Professional Documents
Culture Documents
R esearch
T ools in
A ncient
N ear
E astern
Studies
JOHN L. HAYES
AND TEXTS
e UNDENA
PUBLICA TIONS
Malibu
1990
AIDS AND RESEARCH TOOLS IN ANCIENT NEAR EASTERN STUDIES
All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any
means, electronic or mechanical, i ncluding photocopy, recording, or any information storage and
retrieval system, without permission in writing from the publisher.
Issued under the auspices of I1MAS-The International Institute for Mesopotamian Area Studies
Anyone who has ever tried to learn or to teach Sumerian faces a difficult task. First of
all, knowledge of Sumerian is still at an imperfect stage, with fundamental questions yet to
be resolved. Second, there is a lack of both scholarly and pedagogical tools. Although a
recent descriptive grammar exists, there is no up-to-date sign list or dictionary, and there is
no text-book of any kind. This situation makes it difficult for both student and teacher, and
makes it virtually impossible for someone to learn Sumerian without a teacher.
The aim of this book is to help alleviate this situation. It is a textbook of the Sumerian
language, based on the royal inscriptions of the Ur III period. It is self-contained, so that it
will be of use to students with or without a teacher. It includes a general description of the
Sumerian language and its writing system, and then a series of graduated lessons. Each
lesson contains: sign-list and vocabulary; notes on selected vocabulary; text(s) in cunei
form, either photograph or autograph; transliteration, transcription, and translation; line-by
line commentary on the text. Each lesson concludes with discussions, arranged the
matically, of grammatical issues raised by the text, and of the meaning, function, and
historical context of the text. Later lessons also include supplementary texts for review and
practice, with no new vocabulary or grammar. In each lesson the grammar has generally
been presented inductively from the texts. Finally, there are several appendices, some
treating more general topics, and some serving as reference; the last of these is an index to
grammatical (and other) points.
This book has been designed for a one-semester, three-hour per week class. It can
serve as an introduction to the language for students who will not pursue their study of
Sumerian any further, but it will also prepare students for more advanced work.
Two possible audiences are envisaged. The first is composed of those students who
are comfortable in Akkadian, and who wish to learn Sumerian principally because of their
interest in Mesopotamia. The second is composed of those students who are more
comfortable in West-Semitic, and who wish to learn Sumerian principally because of their
interest in Ebla. The latter audience will either not have studied Akkadian at all, or will
have studied it at some time in the distant past, and may have forgotten much. A certain
amount of material for this latter audience is included which will already be known to those
who are familiar with Akkadian. Throughout, a knowledge of basic linguistic terms and
concepts has been assumed. Since the learning of cuneiform signs often seems like an
onerous chore for those students primarily interested in West-Semitic, the book has been
designed with sufficient emphasis on transliteration and transcription to allow it to be used
without learning the signs.
This book is based on the language of the royal inscriptions of the Ur III period. It is
thus a grammar solely of the written form of the language. It attempts to be purely
synchronic, avoiding a mixture of synchronic and diachronic levels. At the same time,
areas of disagreement about the language are pointed out. Some stress has been placed on
the methodological principles involved in studying a language like Sumerian. Since many
of the problems in understanding Sumerian phonology, morphology, and even syntax are
iii
iv Preface
rooted in difficulties with the script, a certain emphasis has been placed on the nature of the
Sumeriim writing system.
In order to give an idea of the context in which the texts are rooted, some
archaeological, historical, and cultural information is included. Similarly, typological
observations about the Sumerian language have been pointed out, to show that there are
other languages which work in ways similar to Sumerian.
Because of the limited subject-matter of the texts which are used here, not all features
of the language are encountered. Some of these features are touched upon in Lesson 23,
where some alternative views of Sumerian grammar are sketched. Appendix 5 discusses
the ways by which students, including those working alone, can deepen their understanding
of Sumerian. This book will be followed by a second volume, consisting of heavily
annotated extracts from Inanna's Descent. The reading of a major literary text will intro
duce students to a number of problems not encountered in reading the rather stereotyped
texts used in this book.
Appendix 4 is a basic bibliography of the most important and interesting books and
articles on Sumerian. In order for students to become become acquainted with the names of
some of the scholars in the field, a number of modern-day Assyriologists and Sumerolo
gists are quoted throughout the book; all works so quoted are listed in Appendix 4
The genesis of this book goes back to my teaching of Sumerian at the University of
California at Los Angeles. It is a pleasure to thank those who have helped out along the
way. Thorkild Jacobsen was my first teacher of Sumerian; his influence can easily be seen
throughout the book. Sara Denning-Bolle graciously drew the cuneiform signs used in the
sign-lists and those scattered throughout the book; I am especially grateful to her. Barbara
De Marco made a number of useful stylistic observations, and helped in the overall
structure. Several individuals read earlier gestations; I would especially like to thank Daniel
Foxvog, Samuel Greengus, and Stephen Lieberman. Other individuals read certain sec
tions; I thank Denise Schmandt-Besserat and Russell Schuh. James Platt, who studied
from this book, made a number of suggestions. Christopher Walker helped me attain
access to a number of photos from the British Museum. Giorgio Buccellati helped in many
ways, from the initial conception to the final product. And, I would like to thank the staff at
Undena Publications, especially Frank Comparato and Patricia Oliansky. Faults remaining
are my own; I would be very grateful to hear from readers with suggestions for revisions.
I would like to dedicate this book to my mother, for her support and encouragement
over all the years.
CONTENT S
PREFACE . . . . . . . . . i
INTRODUCTION
Importance of Sumerian . . . . . 1
Difficulties in studying Sumerian . . 1
Historical background and texts used 2
CLASSIFICATION . . 5
Linguistic affiliation 5
Dialects . . . . . 5
Typological characteristics 7
Ergativity . . 7
Agglutination . 10
WRITING SYSTEM . 11
External characteristics 11
Original nature . . 12
Internal principles 13
Transliteration 14
Transcription 16
PHONOLOGY 18
Problems . 18
Vowels . . 19
Consonants 20
Other features 23
Lesson 1 25
Lesson 2 47
Lesson 3 57
Lesson 4 65
Lesson 5 73
v
vi Contents
Lesson 6 79
Lesson 7 85
Lesson 8 95
Lesson 9 101
Lesson 1 0 . 1 09
Lesson 1 1 . 1 17
Lesson 1 2 . 129
Lesson 1 3 . 147
Lesson 1 4 . 157
Lesson 15 . 1 65
Lesson 1 6 . 181
Lesson 17 . 191
Lesson 1 8 . 205
Lesson 19 . 21 1
Lesson 20 . 223
Lesson 2 1 . 23 1
Lesson 22 . 245
Lesson 23 . 257
finponBnce of S wnerian
Assyriologists that there are as many Sumerian languages as there are Sumerologists"
(1976:99). Similarly, ThorkildJacobsen has recently said:
Knowledge of Sumerian is still in a rudimentary, experimental stage where
scholars differ on essential points, so that translations, even by highly
competent scholars, may diverge so much that one would never guess that
they rendered the same text. ... Scholars have not yet been able to agree on
basic gramma r and its restraints (1987:xv).
In certain ways, however, it is actually easier to study Sumerian than it is to study, for
example, Akkadian. This is because Sumerian does not have (at least, there is not visible) a
great deal of "morphology"; there are not a large number of grammatical forms to learn.
There is nothing like the weak-verb systems of Akkadian and Hebrew, which require a
great deal of sheer memorization. Rather, many students find the difficulties to be more
conceptual in nature: the language works in ways different than English, or other
languages which students are likely to have been exposed to. It is sometimes difficult to
understand some of these principles, and even more difficult to observe these principles in
action.
The texts utilized here are all royal inscriptions of the Ur III Dynasty (approximately
2112-2004 BC), sometimes referred to as the Neo-Sumerian Dynasty. It grew out of the
vacuum left by the collapse of the Dynasty of Akkad, which had been ruled by Akkadian
speaking kings of Semitic stock (approximately 2334-2193 BC).
The Ur III Dynasty was founded by Ur-Nammu, who ruled in the city of Ur from
about 2112 to 2095. He had previously been governor of Ur under the suzerainty of the
king of Uruk, Utu-Hengal; he may have been a relative of the latter. At some point he
declared himself independent. During his rule, and especially during the rule of his son
Shulgi, the territory controlled by Ur expanded, until it reached most of the area previously
controlled by the rulers of Akkad, that is, most of central and southern Mesopotamia. After
three more descendants of Ur-Nammu, the dynasty collapsed in 2004, partially due to
pressures from the intrusion of nomadic, Semitic-speaking tribes. Thus, the Ur III period
lasted a little more than a century; with the fall of Ur, Sumerian civilization, for all intents
and purposes, also fell.
Ur III was a period of relative calm and stability in much of Mesopotamia. Because of
the blooming of Sumerian art and literature, which had been somewhat submerged under
the Semitic dynasty of Akkad, this period is often called the "Sumerian Renaissance".
Towns were fortified, many temples were rebuilt, and canals were dredged; trade with
various foreign countries flourished.
The city of Ur itself, the capital of the Ur III Dynasty, was primarily excavated by Sir
Leonard Woolley, perhaps the most famous of all Near Eastern archaeologists. The
principal results were published by him and others in a series entitled Ur Excavations. Ten
volumes have appeared: Volume I in 1929, and Volume VII in 1976 (Volume X appeared
in 1951). Woolley popularized his results in a one-volume work entitled Ur of the
Introduction 3
Chaldees (1929). After Woolley's death, P.R.S. Moorey revised and updated the work; it
appeared as Ur 'of the Chaldees' (1982). This is a readable and interesting description of
the city at different historical periods.
Many Ur III texts have been preserved. The vast majority are economic and
administrative; these number in the tens of thousands. Unfortunately, there are very few
texts of what might be called a "historical" nature. There is much that is not known about
such matters as Ur-Nammu's rise to power, the internal politics of the Ur III Dynasty, or
even the physical extent of the Ur III "Empire"; C. J. Gadd refers to the "tantalizing want
of information due to the singular unwillingness of the age to record even the triumphs,
much less the failures, of its kings" (1971:617).
Some original literary texts are also preserved from this period, as well as older works
now committed to writing. Jacobsen says that the kings of Ur Ill, especially Shulgi,
were much concerned to preserve extant older literary works and to
encourage the creation of new ones. The court background of these works is
unmistakable. ... A major portion of Sumerian Literature as we have it traces
back to the court of the kings of the Third Dynasty of Ur, where it was
composed and performed by the royal bards (1987:xii, 277).
The royal inscriptions of the Ur III kings have been the object of study by W. W.
Hallo. According to Hallo's definition, royal inscriptions are texts which "were dedicated
either by, or to, or on behalf of the king" (1962:1). Hallo catalogued these texts, providing
a standard system of reference. He also studied the different sub-types of royal
inscriptions, categorizing them according to their function and according to their form.
These texts range in difficulty, from quite simple to very complex. They also contain a
high degree of formulaity; many of the epithets of the king, for example, occur in a large
number of the inscriptions. Even the phrasing of the verbal expressions is rather fixed.
Since the genre of royal inscriptions existed both before and after the time of Ur III (in
Sumerian and in Akkadian), a knowledge of the Ur III texts gives immediate access to
other similar texts.
There has been much recent discussion about when Sumerian ceased to be a spoken
language. This is not an easy question to answer; there are both historical issues and issues
of general linguistics to resolve. (The subject is further discussed in Appendix 1.) Most
Sumerologists would say that Sumerian was a living spoken language in Sumer during the
Ur III period, although some would say that it was already starting to die out during the
latter part of this period. A minority would say that spoken Sumerian was either pretty far
on its road to extinction, or might even have ceased to be a spoken language by the end of
the Ur III period. Even the proponents of this view, however, would admit that the
language of the Ur III royal inscriptions is "good" Sumerian, unlike some Sumerian of later
periods.
PART ONE : THESUME IDAN LAN GUA GE
Dialects
The S umerians referred to their own language by a term often transliterated as: eme
gir 15. The value of the second sign is not sure, and so the term is variously transliterated as
eme-gi7' eme-lrn, etc., especially in older secondary literature. erne means "tongue" in
Sumerian. The meaning of girl S is unsure. Older scholars thought that it meant "Sumer";
in that case, the term would mean "language of Sumer". More recently it has been argued
that the term means something like "noble, prince"; erne-girl S would then mean "the noble
language". Because of the uncertainties in reading this word, the term "Main Dialect" is
often used instead.
There is also a "dialect" called erne-sal. The meaning of the second element of the
name is uncertain; it may mean "fine, thin". The "status" of this dialect is also uncertain. It
has traditionally been called a "women's language", because it appears in literary texts of the
Old B abylonian period, used by women when speaking to other women. For example, in
the myth "Inanna's Descent to the Netherworld", when Inanna speaks to her aide Nin-
5
6 Manual ofSumerian
Shubur, she does so in Emesal. There is no consistency in this usage; in other texts Inanna
may speak in Main Dialect. Moreover, in texts of the later Old Babylonian period Emesal
is also used for specific genres of text. Certain kinds of lamentations are always written in
Emesal, even though recited by male priests. (Texts in some of these genres were
preserved and even composed in schools for a thousand years after Sumerian had ceased to
be a spoken language.) This use by men makes it difficult to determine exactly what
Emesal is, and whether or not it should be classified as a "dialect".
Emesal is well-attested from the beginning of the Old Babylonian period on.
However, there appear to be at least one or two Emesal forms in the Gudea texts, and there
has been a recent attempt to see Emesal forms in a group of texts written in an unusual
orthography from Tell Abu Salabikh (approximately 2600 BC).
Emesal differs from Main Dialect in phonology and in the lexicon, but not apparently
in morphology. In phonology, the Emesal forms often appear to be older. For example,
the word for "lord" in Main Dialect is len/, in Emesal lumun/. It is difficult to say exactly
what the more original form was; it may have been something like */ewenl or */uwun/.
In any case, the Emesal form appears more conservative than the Main Dialect form.
According to other scholars, however, Emesal forms are linguistically the more innovative;
Emesal forms result from consonants being shifted to a more fronted or to a higher place of
articulation. For example, Main Dialect Igl > Emesal Ib/; Main Dialect Idl > Emesal Iz/,
etc. But there are several exceptions to these general principles, and there are a number of
details of Emesal phonology which are not clear. As an example from the lexicon, the
Main Dialect word for the interrogative "what?" is lana/; the Emesal form is /tal. These
are apparently two etymologically distinct words.
It has been claimed that Emesal shares certain characteristics of "women' s languages"
which occur elsewhere in the world. In particular, women's languages are said to differ
from "standard" dialects in phonology - the women's dialect being more conservative than
the standard dialect - and in the lexicon. More work needs to be done in defining the
characteristics of Emesal, and in comparing Emesal with other women's languages.
Not much is known about geographical variation within Sumerian. The extent of the
Sumerian-speaking area i unsure; Sumerian texts are preserved from only a rather limited
area. Moreover, the nature of the Sumerian writing system makes it difficult to see such
vanatlOn. Only traces can be found, particularly in the later periods. There was
undoubtedly more dialectal variation present than the writing system allows us to see.
Similarly, although Sumerian was spoken over a long period of time, there does not
appear to be much variation before the Old Babylonian period. More differentiation is
noticeable in post-Old B abylonian periods, when Sumerian was no longer a spoken
language. But here the differences may reflect the practices of different scribal schools and
scribal centers, and not differences which were originally in spoken Sumerian.
There are occasional references in late Sumerian texts to what are apparently
specialized languages, or jargons of particular occupations. For example, there are passing
references to eme-utula, "the language of shepherds", and to eme-ma-Iah4-' "the language
of sailors". It is hard to say what these dialects or jargons were like. S imilarly, there are
only passing references to what may be some kind of "literary dialects": erne-gal, "great
Classification 7
language", eme-sukud, "high language", etc. It is not known what these designations
mean.
T ypological characteristics
There are several ways in which Sumerian works differently than the Semitic or Indo
European languages. Consider the Akkadian sentence, "The king went":
Now, consider the Akkadian sentence, "The king built the house":
In (3), the subject of the intransitive verb is marked by .0, the absolute case-marker.
In (4), the subject of the transitive verb is marked by .e, the ergative case-marker, while the
direct object is marked by .0, the absolute case-marker. This fits the definition of an
ergative language: The subject of a transitive verb is marked one way (in Sumerian, by .e),
while the subject of an intransitive verb, and the direct object of a transitive verb, are
marked a different way (in Sumerian, by .0).
Ergativity is a different way of marking the primary participants in a sentence. In an
accusative language, the subject of a transitive verb and the subject of an intransitive verb
fall into one grammatical category; in an ergative language, the subject of an intransitive
verb and the object of a transitive verb fall into one grammatical category. Consider the two
English sentences, "The ball rolled down the hill", and "The boy rolled the ball down the
hill". In English, "ball" in the first sentence is the subject, but in the second sentence it's the
direct object. Yet in each case, it's the ball that is rolling down the hill. In an ergative
language, "ball" would be in the absolute case in both the first and second sentences, and
"boy" would be in the ergative case in the second sentence. In this example, an ergative
language seems to capture our intuitions about the role of the ball in these two sentences
better than does our accusative language.
In the above discussion, the terms "subject" and "object" were used. However, it is
imprecise (and unjustified on theoretical grounds) to use these two terms when talking
about an ergative language. Most linguists prefer to use the term "agent" to refer to the
subject of the transitive verb (marked by the ergative case), and the term "patient" to refer
both to the subject of the intransitive verb, and to the direct object of a transitive verb (both
marked by the absolute case). Thus, in the examples above, "boy" is the agent, and "ball"
is the patient. In practice, it is very difficult to escape using such common terms as
"subject" and "object", especially in unambiguous contexts, even if these terms do not
really fit Sumerian.
There are many ergative languages in the world, belonging to a number of different
language families: many languages in Australia, many American Indian languages, the
Caucasoid languages (for example, Georgian), Basque, to name a few. However, none of
what are sometimes referred to as the "major cultural languages" of Europe are ergative,
and so the concept is unfamiliar.
There are two other important points about ergativity. First, the definition given above
describes what may be called "minimally" ergative languages. However, ergativity can also
be reflected in other parts of a language's grammatical system - it may affect verbal
agreement, cross-referencing of case-markers, coordination and subordination, etc. This
will be discussed in more detail later.
Second, there appear to be very few (if any) "pure" ergative languages. Most (perhaps
all) ergative languages are "split". In certain constructions, the language behaves in an
ergative manner; in other constructions, the language behaves in an accusative manner. In
S umerian, for instance, the perfect aspect functions in an ergative manner, while the
imperfect aspect functions in an accusative manner. That is, Sumerian is split along an
aspectual axis. There are other languages in the world which are split along exactly such an
axis, that is, the perfect aspect functions in an ergative manner, and the imperfect aspect
Classification 9
Diakonoff ( 1965) sketched the system of ergativity in Sumerian and other Ancient Near
Eastern languages, without explaining the details of morphology. The articles by Daniel
Foxvog ( 1 975) and Piotr Michalowski ( 1980a) viewed Sumerian in an explicitly ergative
framework, while elucidating the verbal morphology. Van Aalderen ( 1 9 82) has explored
some of the theoretical issues in more detail. The grammar by Marie-Louise Thomsen
( 198 4) also follows a split-ergative model. A recent survey of ergativity in Sumerian is by
Gong Yushu ( 1987).
Agglutination
Sumerian is often described as an "agglutinative" language. This term goes back to the
nineteenth century, when linguists attempted to classify the languages of the world into a
few basic types, based solely on typological (not genetic) criteria. For these linguists, the
three most common types of language could be classified as:
Isolating
In isolating languages, virtually every morpheme forms a separate "word". In
Chinese, for example, there are no tense-markers on verbs; such information is conveyed
by separate adverbs. There are also no plural-markers on nouns or verbs; this information
is conveyed by separate number-words.
Fusional
In fusional languages, such as Akkadian or Latin, grammatical morphemes are
expressed through endings on nouns or verbs, and several different morphemes tend to
"fuse" together. Latin amo, for example, means "I love". The /0/ ending on the verb
signals several things: the verb is first person, singular, present tense, indicative mood,
active voice. However, none of the morphemes for person, number, tense, mood, or voice
can be segmented out - they are all fused into the ending /0/.
Agglutinative
In agglutinative languages, as in fusional languages, several grammatical morphemes
are combined into one word. However, the morphemes are distinct from each other; they
do not fuse together. In an agglutinative language, strings of prefixes or suffixes tend to
occur; each affix is formally distinct, and expresses one morpheme. The parade example of
a language of this type is Turkish. In Turkish, the phrase "from his houses" is expressed
as: evlerinden. Ev means "house", ler is the plural marker, in is the possessive pronoun
"his", and den is the postposition expressing the ablative "from". In general, each affix
expresses one morpheme; each morpheme is invariant: ler is the automatic plural marker
for all nouns; den means "from" after any nominal phrase, etc. The morphemes are distinct,
not fused into each other.
Sumerian is similar to Turkish. The verbal phrase, for example, consists of a string of
prefixes, followed by the verbal root, and then a smaller string of suffixes. Each affix
expresses one morpheme, and each affix is (basically) invariant. Nominal phrases can be
very long, with a noun, modifying adjectives and appositives, genitive phrases, etc., with a
Writing system 11
WRITING SYSTEM
External characteristics
In discussing any writing system, there are two factors to consider: the external
characteristics of the script, and the principles behind the script.
Because of the external shape of the signs in the Sumerian script, its writing system is
called "cuneiform". "Cuneus" is the Latin word for "wedge"; the term was coined because
of the most striking characteristic of the script - the fact that the signs are built up of strokes
looking like little wedges. (The term cuneiform was apparently first used by one Thomas
Hyde in 1700. In his Historia religionis veterum Persarum, he refers to "dactuli pyra
midales seu cuneiformei".)
The cuneiform signs were inscribed by means of a stylus probably formed from an
actual reed (such as still grows in modern-day Iraq), by impressing the stylus upon a tablet
of moist clay (or, occasionally, upon other surfaces). The stylus could also be made of
bone, metal, hardwood, or even other material.
The first cuneiform texts discovered were all relatively late, from a period when the
wedge-shaped characteristics of the script were most striking. In the earliest phases of the
script, however, this wedge-shaped character is less pronounced; the script of most of the
Ur III inscriptions in this book does not look nearly as wedge-shaped as do later texts.
12 Manual ofSumerian
The tenn cuneifonn refers solely to the external shape of the individual signs.
Cuneifonn script was adopted and modified by many peoples of the Ancient Near East; it
was used to write Akkadian, U garitic, Hurrian, Persian, etc. However, the fact that these
languages use signs with the same general external characteristics says nothing about their
possible genetic relationship. Sumerian, Akkadian, Hurrian, and Persian, for example,
belong to four entirely unrelated language families. Expressions such as "cuneifonn
language" are occasionally encountered, but this is a rather imprecise way of referring to
one or several languages, which may or may not be related, which use a script with the
same external characteristics.
Original nature
The writing system used for English is an attempt to render speech as closely as
possible. Although English does suffer from numerous archaic spellings, and there are
certain features (such as upper and lower-case letters) which are found only in writing,
writing is basically an attempt to reproduce speech sounds. By contrast, the Sumerian
writing system was never an exact, phonetic representation of speech; it was not "designed"
to reproduce spoken language as such. Rather, to some degree the writing system is only a
mnemonic device, to jog the memory of the writer and reader. The earliest uses of writing
were for administrative texts, which were of a fonnulaic nature, and whose contents were
familiar to the scribes. There was no need to write down what would be obvious to a
scribe who was a native speaker of Sumerian, and who was familiar with the material being
written. When such scribes "read" the texts, they knew how to supply the infonnation not
indicated explicitly in the writing.
Thus, a certain amount of infonnation in the spoken language was not expressed in the
writing. The further back in time one goes, the less the Sumerian writing system expresses
grammatical elements which are assumed to have been present in the spoken language. For
example, the basic graphic shape representing the root for "to build" was originally a
picture of a wooden peg. In the earliest Sumerian, this one sign could be used for any
inflected fonn of the verb: any tense, mood, or person. S imilarly, the expression for "on
that day" in Sumerian was: ud-bi- ("day-that-on"). But in the earliest Sumerian, only the
ud-sign was written; the reader inferred the rest.
As might be imagined, this lack of explicitness in the script can cause much trouble in
interpreting Sumerian texts. Nor is this problem limited to the earliest Sumerian texts; in
late economic texts, for instance, it is often difficult to tell if something is being distributed
"to" or "from" somebody.
As time passed, the scribes wrote more and more down, that is, the writing became
more and more explicit. For example, there is a Sumerian text known as the "Kesh Temple
Hymn", attested in several copies mostly from the Old Babylonian period (dating to around
1 800 BC). In the 1960s, a version of the same text was found at Tell Abu Salabikh, dating
to about perhaps,2500 BC. Unfortunately, only a few lines of the Tell Abu Salabikh
version survive. But if one compares the Old Babylonian version with the Tell Abu
Salabikh version, it can be seen that although the text itself is relatively stable, the Old
Babylonian version indicates more verbal affixes than does the Tell Abu Salabikh version.
Writing system 13
This increase in explicitness may be connected with the fact that Sumerian was gradually
dying out, and so scribes needed more help in their own understanding of texts.
Thus, a fundamental feature of the Sumerian writing system is its lack of explicitness.
It does not fully represent the spoken language. This has been summarized by Jacobsen:
"The history of Sumerian writing is one of progressively ever greater but never quite
attained adjustment to Sumerian speech" ( 1 957:366 n.1). Similarly, Marvin Powell has
pointed out that "We find traces of its mnemonic character enduring to the very end of the
Sumerian orthographic tradition" (1981 :42 1 ).
A further complicating problem is that the writing system is to some degree
morpheme-bound. There is indirect evidence to show that there were certain phonological
changes which took place in Sumerian, such as contraction, vowel deletion, etc., but these
changes are masked by the script; the script often reproduces the basic morpheme, without
showing the changes which are assumed to have taken place in the spoken language.
The view here presented, that the Sumerian writing system in origin and in practice is
basically mnemonic, has been especially expounded by Diakonoff ( 1976) and Stephen
Lieberman ( 1977).
Internal principles
The script used for writing Sumerian is a combination of "logographic" and "syllabic"
elements. Logographic means that a sign stands for a particular word. For example, the
sign 4 stands for the word utu, "sun"; the sign stands for the word digi!:, "god". The
external shape of many of these signs is clearly pictographic in origin. Thus the sign for
"sun" was originally a picture of the sun rising over a mountain. The sign for "god" was
originally a picture of a star. The original significance of many signs cannot yet be
determined.
The same sign can often have more than one logographic value. Thus, the same sign
can represent diEi!:, "god", or it can represent an, "sky". In general, it is only the context
which determines the meaning of the sign, and its correct reading.
Syllabic signs are used to reproduce a sequence of phonetic elements. For example,
the sign is used to represent the syllable Iga/. This particular syllable can form a
component of several different morphemes: it may be part of the cohortative prefix on
verbs, or part of the ending of a genitive phrase on nouns, etc. The sign in these
contexts does not stand for any particular word; rather, its purpose is to represent the
phonetic sequence I gl -I ai, which may form part of a number of different morphemes.
Syllabic signs can represent several different kinds of segments of consonants and
vowels. Some syllabic signs stand for single vowels, e.g., and i. More common are
signs standing for the sequence consonant-vowel (ba, mu) or vowel-consonant (fill, in).
There are some signs that stand for consonant-vowel-consonant, but these are not common;
instead, the script uses a convention that represents ICVCI by CV-Vc. For example, the
segment Inirl is written by: ni-ir. A writing such as ni-ir does not imply a long vowel;
this is purely an orthographic convention, to reduce the number of CVC-signs which
would otherwise be necessary.
Many signs have more than one syllabic value. Many signs have both logographic
14 Manual ofSumerian
and syllabic values - sometimes more than one of each. The correct value of the sign can
usually only be derived from the context. Signs with more than one value are called
"polyvalent", or are said to have several "readings".
Thus, the Sumerian writing system is both logographic and syllabic. The syllabic
value of most signs derives from a logographic value. For example, the sign in its
meaning as "sky" is pronounced lan/. This phonetic value was then generalized, so that
this sign can stand for the syllable lanl in other contexts.
In general, lexical morphemes are written logographically, and grammatical
morphemes are written syllabically, but this is not always the case. The system is
complicated by the fact that certain syllabic signs tend to be used for certain morphemes.
For example, there is a "conjugation-prefix" on the verb, pronounced Ibi!. There are
several different possible ways that this phonetic sequence could be represented in the
script. In practice, however, the scribes almost always used only one of these possibilities,
the sign . That is, certain morphemes tend to be indicated in only one way, and,
conversely, certain signs tend to be used only for certain morphemes.
In addition to logographic and syllabic signs, there are a few other elements present in
the script. One of these is "determinatives". Determinatives are signs which are used to
indicate the general semantic class to which a following (occasionally a preceding) noun
belongs. For example, almost all divine names are preceded by the sign ; this sign tells
the scribe that "what follows is a divine name". Most names of countries are followed by
the sign -+; this sign tells the scribe that "what precedes is the name of a country".
Determinatives were probably not spoken, even when Sumerian was read out loud. They
were only a feature of the written language.
In other contexts, the cuneiform signs which function as determinatives can also
function as logographic or syllabic elements. For example, the sign can represent digir,
"god"; the sign can represent ki, "country".
To sum up, Sumerian is mostly logographic, and only partially syllabic. Akkadian, on
the other hand, is mostly syllabic, and only partially logographic. Persian cuneiform is
almost entirely syllabic, and Ugaritic cuneiform is basically alphabetic. In practice, people
sometimes confuse the issue, and the term cuneiform is occasionally used to refer in general
to any logographic-syllabic system of writing, but this is wrong; there are many
logographic-syllabic scripts which have existed in the world, which are not cuneiform.
This has been a somewhat simplified discussion of the Sumerian writing system.
There has been much recent discussion about the script, mostly hinging on theoretical
questions, such as the difference between pictographic and logographic, or the degree to
which the script is morpheme-bound.
Transliteration
When citing Sumerian texts, or when discussing Sumerian grammar or vocabulary,
Sumerologists do not generally reproduce the original cuneiform signs. Rather, they cite
the word or passage in transliteration into Latin characters. Transliteration is a sign-by-sign
image of the original written text. It is designed specifically to reflect the actual cuneiform
signs present. By looking at a transliteration, one should be able to determine exactly
Writingsystem 15
which cuneiform signs occur in the original text (excluding palaeographic niceties).
Transliteration serves several purposes. It is more convenient, quicker, and cheaper to
produce Latin characters than it is to produce cuneiform characters. Also, it provides an
approximate phonetic rendering of the signs occurring in the Sumerian. Since many
Sumerian signs have more than one reading, a scholar, by giving the text in transliteration,
explicitly states his opinion about the reading of a particular cuneiform sign. For example,
the sign can be read Hkur (the name of a god), or im ("wind"), or ni ("self ' ). Based
on his understanding of the text, a scholar decides the correct reading.
There are some complexities of transliteration. It is possible for several different
cuneiform signs to have the same pronunciation. These signs must be differentiated in
transliteration, so that the original cuneiform can be reconstructed from the transliteration.
For example, there are at least four different signs pronounced as /u/. If y were used as
the transliteration for all four signs, it would not be possible to go backward from the
transliteration: Given a transliteration y, one could not tell which of the four possible signs
actually was written in the cuneiform. To obviate this problem, scholars have devised the
following system: The most common (or most important) sign with a particular value is
unmarked. The second most common (or most important) sign with this same value is
marked with an acute accent: g. The third most common (or most important) sign with this
same value is marked with a grave accent: !l. The fourth, and higher, most common signs
with this same value are marked with subscripts: .!4 , liS ' etc. This system is purely
arbitrary; it provides a convenient means to differentiate between signs pronounced alike,
thus enabling us to reconstruct the cuneiform from the transliteration.
This use of the acute and grave accent-marks as "indices" has nothing to do with
pronunciation. They do not indicate anything about accent, nor do they indicate anything
about vocalic length, nor do they indicate anything about tone. They are used instead of a
possible Y2 and Y3 simply because it is easier to type accent marks (at least in Europe) than
it is to turn the typewriter carriage up to make a subscript.
These indices are based largely on frequency. However, these frequencies were
determined on the basis of Akkadian texts, not on the basis of Sumerian texts (for the
simple reason that Akkadian was "discovered" before Sumerian). This produces a certain
inconsistency. In Sumerian, for example, the bi-sign is much more common than the bi
sign. This inconsistency is not really a problem; the only other alternative would have been
to devise a separate system for Sumerian, based on values and frequencies in Sumerian.
But this would have engendered so much confusion and complication that it is far easier to
work with the traditional system.
Confusion arises when indices are used on bisyllabic signs, that is, signs which
represent a segment of two syllables, such as /kala/ or /Urim/. If there is more than one
sign with the same bisyllabic reading, some scholars put the accent-marks on the first
vowel, then continue onto the second syllable if there are several signs with the same
reading. Other scholars, however, begin with the last vowel, moving back to the first.
Either system is prone to mechanical mistakes in printing, and the mere presence of the two
different systems can cause problems in determining what the cuneiform sign actually was.
To mitigate against this difficulty, some Sumerologists do not use acute or grave accent-
16 Manual ofSumerian
marks on bisyllabic signs. Instead, they use a subscript 2 or subscript 3 when necessary.
For example, there are several signs with the value of /kala/. These are differentiated as:
kala, kala 2> kala3' kala4' etc. This is the system followed here. Some recent publications,
including the Pennsylvania Sumerian Dictionary, use subscripts in place of accent-marks,
even on monosyllabic signs. Thus, instead of !!, they use l!2 ; instead of y, they use l!3 '
Determinatives are a feature of the written language, and were probably not spoken.
To indicate that they were not pronounced, they are transliterated with superscript letters:
Xki, tugX, etc. For convenience sake, the determinative for god (the iligir-sign) is trans
literated as a superscript d : dBtar. Because of the typographical difficulties of printing
superscripts, some publications instead print the determinatives on the same print-line,
connected by a period: X.ki; tug.x.
In transli teration, signs comprising one "word" are linked by hyphens: kalam-ma,
illgir-ra-ni, etc. (Determinatives are an exception; no hyphens are used.) As will be seen
below, it is not always easy to determine what constitutes a "word" in Sumerian.
Some Sumerologists use initial capital letters in their transliteration of Sumerian proper
names; other Sumerologists do not. Those who do not use them, consider capital letters to
be a feature particular to the English writing system; since capital letters have no correlate in
the Sumerian writing system, they should not be used in transliteration. Other scholars feel
that since transliteration is an artificial device anyway, there is no harm in using capital
letters, if they help make the text clearer to the reader. This second practice is followed
here.
Finally, it is necessary to say a few words about the typographic conventions used in
transliterating Sumerian. Throughout this book, Sumerian is transliterated by Roman
characters, underlined. The few Akkadian citations used here follow the same system.
However, it is occasionally inconvenient to use the same typographic conventions for two
different languages. To solve this problem, many publications cite Sumerian in Roman
characters, but widely-spaced. Thus, the word for "god" will be transliterated as: digir.
This may seem like a convenient procedure to differentiate citations from the two
languages, but it is prone to produce mechanical errors in printing.
It is frequently the case that it is not known how a particular Sumerian sign (or word)
is to be read. Some scholars elaborate the system just discussed, by presenting such
doubtful or unsure readings in caps. For example, the word for "interest-bearing loan" in
Sumerian is written: 1lUI . It is not sure how the first sign is to be read. For this
reason, the word is often cited as: IjAR-ra. Some scholars do, however, believe that they
now know how to read this word, and so nowadays one is likely to see the reading: urS
ra. That is, wide-spaced Roman is used for the "standard" transliteration of Sumerian, and
caps Roman is used for unsure readings. Not all Sumerologists follow this system,
however, and what is sure for one scholar may be unsure for another scholar.
Transcription
Transliteration is, by definition, a reflection of the written language, and so does not
necessarily reproduce the spoken language well (as we think we understand it). For this
reason, most Sumerologists use some form of transcription in their study of Sumerian.
Writingsystem 17
example, could be translated without knowing much Sumerian, simply from a knowledge
of Akkadian and of simple vocabulary; a transcription reflects the structure of the language
hidden beneath the written form.
At certain times in this book, the purely phonemic structure of Sumerian will be
stressed, ignoring any morphological considerations. In that case, normal linguistic practice
will be followed, and the item will be put between slashes, e.g., Ikalama/.
Thus, our understanding of Sumerian may be reflected in three different ways: a
transliteration, reflecting the written shape; a phonemic transcription, reflecting the
pronunciation; and a morphological transcription, reflecting our understanding of the
pronunciation and morphology.
PHONOLO GY
Problems
Vowels
Sumerian had at least the following vowels:
i u
e
a
The precise phonetic value of these vowels, particularly the I el, is unsure.
Many scholars also believe that Sumerian had an lol-quality vowel, but since no 101
existed in Akkadian (at least on the phonemic level), there is only indirect evidence to
reconstruct it. It is very difficult to determine whether any particular Sumerian word had an
101 -quality vowel or an lul -quality vowel; its existence has been established for only a few
cases. Under the assumption of the existence of this 101 -quality vowel, the vocalic system
of Sumerian is more symmetrical:
i u
e 0
a
Other Sumerologists have posited other vowels, such as both an open lel and a closed
le/. Others have posited the existence of nasalized vowels, but the exact number and
quality of these varies from one scholar to another: /i/; le/; /i I and la/; /i I, lal and lel,
etc. Claude Boisson ( 1988) has investigated various reconstructions of the phonemic
system of Sumerian, in comparison with what is known about language in general. He
feels that if Sumerian possessed only four vowels, then the vowel normally represented as
20 Manual ofSumerian
lel was more likely IEI than le/. He also feels that none of the systems of nasals which
have been posited for Sumerian is likely.
It is not sure if there was a phonemic distinction between short and long vowels; this
cannot be told from the script. It has been postulated that there were no originally long
vowels in Sumerian, but that they did arise through vocalic contraction, in particular the
contraction of final root-vowels with initial vowels of suffixes.
As discussed above, in practical terms most transliterations of Sumerian usually only
reflect the vowels known from Akkadian; that is, the four vowels listed above.
Consonants
Most analyses of Sumerian would include the following consonants:
b p m
d t n
g k IJ
z s
b
1 r
(For ease in printing, the consonant indicated above as b is often simply transliterated
as h, without the "dish". Since Sumerian does not have a "simple" IhI, there is no
ambiguity in this usage.)
Virtually all Sumerologists accept the existence of the velar nasal IIJI (although some
scholars prefer to speak of a palatal nasal, and others have seen more complex phonemes,
such as IIJm/). When Sumerian words containing this phoneme are loaned into Akkadian,
it is usually (although not always) reflected as !!g. For example, , "kind of priest"
(Lesson 2 1 ) appears in Akkadian as ang'y.
Transliterations of this phoneme vary. In older works, and in many contemporary
works, it may simply appear as g. Some recent works use g, or some typographical
equivalent (g, etc.). It will be transliterated here as g, in cases where it is assumed by most
Sumerologists to be present. With many words, however, it is not known whether a
phoneme is IIJ /, Igl, or even 1nl or Iml, and so some variation in the transliteration of
certain words appears. For example, the verb "to go" is understood by some
Sumerologists to be Iginl, but by others to be Iginl (or Igen/).
Many Sumerologists believe that Sumerian had a phoneme usually symbolized by
Idr/; its exact phonetic significance is unsure. Its existence has been proven in only a few
cases. Because of the difficulties of proving its existence in specific words, it is usually not
indicated in transcription; instead, in the standard sign-lists and in most transcriptions it is
reflected as g.
Several other consonants have been posited for Sumerian: Ihl, Iwl, Iy/; two (or
more) types of /11; two (or more) types of Ir/; a labiovelar Ikw/; a pre-nasalized labial stop
I mb/; etc. Since none of these sounds exists in Akkadian, the evidence for their existence
in Sumerian is indirect at best, and individual Sumerologists have their own preferences.
Phonology 21
e i b p m
a 0 d t n z
u g k g
z s
b
1 r f
In the tables above, certain consonants are indicated as differing only in voice: Ib/
Ip/; Idl - It/; etc. It is not in fact sure what differentiated such pairs; Lieberman explicitly
says that the distinction he marks as Ibl - Ipl was not one of voice. Some Sumerologists
have speculated that the difference was one of aspiration; this is not an uncommon view
today. Boisson, for example, says: "A correlation of aspiration seems to be the only
hypothesis with a high probability of success" ( 1988:25). Other Sumerologists have
speculated that the difference was one of glottalization.
There does not appear to have been a phonemic distinction between short and long
consonants; it is not in fact sure if long consonants occurred at all.
One of the thorniest questions in Sumerian involves the status of word-final and
syllable-final consonants. According to most Sumerologists, certain consonants, when in
word-final position, were not pronounced. For example, the root for "dais" is Ibarag/,
with a word-final Ig/. However, unless this Igl was followed by a vowel, it was not
pronounced: this word would have been pronounced as Ibara/.
The word-final consonant in a root is usually referred to by the German term
"Auslaut". Thus, it is said that the word for "dais" (pronounced Ibarag/) had a "g
Auslaut", or the word for "to live" (pronounced Itil/) had a "l-Auslaut".
The consonants which were regularly not pronounced in word-final position are called
"amissable" consonants. Those which were pronounced in word-final position are called
"non-amissable". (These terms are apparently peculiar to Sumerologists; they are not used
by general linguists.)
Sumerologists differ among themselves about which consonants were not
pronounced. Some believe this affected all consonants, although perhaps not "to the same
degree". Others believe that it affected a smaller number of consonants (although no two
lists of such consonants seem to agree exactly). Also, it is not known if the amissable
consonants were not pronounced in word-final position only; most Sumerologists believe
that they were not pronounced in any syllable-final position. Arno Poebel, for example (the
real father of Sumerian grammar), states that "As a rule, an amissable consonant is dropped
whenever it stands at the end of a word or syllable" ( 1935: 147). Similarly, Samuel Noah
Kramer says: "All final consonants in Sumerian are amissable. ... The term 'final
consonant' as here used includes the consonant at the end of a syllable as well as the one at
the end of a word" ( 1 936: 1 9).
22 ManualofSumerian
as ti(1), others as ti(1). In this book, all word-final consonants have been consistently
transliterated (and transcribed) .
Other features
There were undoubtedly other features in the spoken language, which the writing
system only hints at. There is only marginal evidence, for example, to determine word
stress, and it will not be dealt with here. Similarly, there is only the most indirect evidence
for sentence-intonation.
Because of what is claimed to be a large number of homonyms in Sumerian, it has
several times been argued that Sumerian possessed phonemic tones. Diakonoff, for
example, says: "Sumerian was certainly a tonal language, or else the many homonyms
would have made spoken Sumerian quite unintelligible" ( 1983:86). However, the evidence
is indirect and slight. In fact. many words which earlier Sumerologists believed to be
homonyms have been shown to contain different Auslauts, and so are not actually
homonyms.
PART lWO: LESSONS IN SUMERIAN GRAMMAR
Lesson 1
This first text is a royal inscription of Ur-Nammu, the founder of the Ur III Dynasty
(ruled 2 1 12-2095 BC).
an heaven
M house
p dti to build
4 na
Tt !
P ni
25
26 Lesson 1
}-J!-T ma
Jm ke4
.( mu
.(
Especially for those who are primarily interested in West-Semitic, it is not always easy
to master cuneiform signs. In certain ways, however, it is easier to learn the signs of this
period than the signs of later periods. In later periods, the repertoire of possible sign
shapes becomes quite reduced, so that (superficially) the signs of the Neo-Assyrian period,
for example, all look very similar. In the earlier periods, however, the signs are much more
distinctive, making them easier to learn.
However, one problem in studying the signs of the early periods is the occasional
wide variation in external shape of the signs. For example, the sign for , "house", looks
rather different in Text 2 than it does in Text 1 . This variation is due to several factors:
nature of the writing surface, different scribal traditions at different scribal centers,
individual idiosyncrasies of handwriting, etc. The sign-lists and vocabularies attempt to
produce the basic or essential shape of each sign; the signs in the autographs are
reproduced exactly as published.
Notes
The Notes discuss some of the more important vocabulary items. Often, reference is
made to Akkadian words which were borrowed from these Sumerian words. This practice
is open to methodological criticism, since Akkadian is not Sumerian, and there is no reason
to assume that Sumerian words always kept exactly the same meaning when placed into an
Akkadian context. But since normally much more is known about the Akkadian term than
about the Sumerian term, it is still useful to examine the Akkadian equivalents.
Nanna The city-god of Ur. The large temple-complex at Ur discussed below was sacred
to him in particular. He was associated with the moon; nanna in fact means "moon".
In Akkadian, the word nannaru occurs, glossed by the CAD as: "luminary, light (as
poetic term, an epithet of the moon god and !Star)". This Akkadian word may be some
kind of blend or contamination between the Sumerian word nanna and the Akkadian root
nawanI.
Because of this Akkadian word, some earlier Sumerologists believed that the
Sumerian word had an Ir/-Auslaut, and so the name sometimes appears as Nannar.
However, there seems to be no inner-Sumerian evidence which would indicate such an
Auslaut.
The moon-god was also referred to as Zuen; this problem will be further discussed in
Lesson 1 3.
The Mesopotamian scribes interpreted the cuneiform sign expressing his name as
J:
consisting of two signs: the SeS-sign ( or }> ) followed b the ki-sign ).
Therefore, in older works the name is sometimes transliterated as: SeS-ki. More likely,
Lesson 1 27
however, the second element was originally the na-sign, functioning as a phonetic
complement of some kind.
Nammu Not much is know about this goddess. However, she is described as "the mo
ther who gave birth to heaven and earth", and as "the primeval mother, who gave birth to all
the gods". It is thus possible that at one time she played a more important role in Sumerian
cosmogony.
The cuneiform sign which represents this name can also be read engur, which lexical
texts equate with the Akkadian apsu, the "watery deep" (see Lesson 14). The cuneiform
sign may be an abstract representation of this deep.
In some older Sumerological works, the two readings of this sign (Nammu and engyr)
were not clearly differentiated. Therefore, the name of the founder of the Ur III Dynasty
sometimes appears as Ur-Engur, or Ur-Gur.
UrimS In English, "Ur". One of the more famous cities in southern Mesopotamia; the city
after which the Ur III period is named. The name of the modern site is al-Muqayyar.
The etymology of the name Urims is unknown. It is also not known how these two
particular cuneiform signs (presumably, the e-sign followed by the ab-sign) came to
represent the name.
Urims is the long value of the sign. The short value is variously transliterated as Uri ,
Uri2, Uri3 , or Uris . The oscillation in diacritics illustrates the problem of diacritical marks
on bisyllabic signs.
The sign-lists in this book give the long value first, followed by the short value.
Because both are encountered in Sumerological literature, it is necessary to know both
values, even though this seems like a totally unnecessary burden upon the student.
Sometimes, the name is written e-unug and not e-ab, in which case it should pro
perly be transliterated as Urim2.
The English equivalent, "Ur", derives from the Old Testament f'ur kasdlmf, "Ur of
the Chaldeans". Exactly how the Hebrew f'urf derives from the Sumerian fUrimf is
unsure.
nin In general, the Sumerian word for "lord" is en; the feminine equivalent, "lady", is nin.
(It is not impossible that the two words are etymologically related.) However, in older
Sumerian nin can also be used to refer to masculine entities. Perhaps at one time the term
was genderless. In the Ur III period, this usage can be considered an archaism.
ur The usual interpretation of this word is something like "man; warrior, hero". In
bilingual lexical texts, ur is glossed as amelu, "man", and as kalbu, "dog". ur with the
meaning "dog" is not uncommon in Sumerian texts. However, ur meaning "man" seems to
occur only in personal names; it does not have this meaning in actual texts (although the
compound ur-sag, "hero", presumably "man-head", is common).
One might guess that the ur-sign was originally a picture of a dog or some kind of
beast, but even the earliest attestations of the ur-sign do not look very animal-like.
l ugal Etymologically, a compound of lu "man" and gal "great". This word is further
28 Lesson 1
discussed in Lesson 7.
According to 1.1. Gelb, "The Sumerian word has several meanings: a) a dwelling
house, even a room b) palace, temple c) family, clan d) household. The same meanings
occur also for the Akkadian bItum" ( 1979b:2). In the sense of "temple", it can refer either
to one particular building, or to an entire temple complex consisting of several buildings.
In very recent secondary literature, it is occasionally transliterated as: 'a.
dd Although du occasionally means "to build" de novo, it more often means "to rebuild".
It is especially frequent when describing the rebuilding of temples which had fallen into
disrepair. Usually, it is difficult to tell in any particular text whether du means "to build" or
"to rebuild"; this can only be resolved by historical or archaeological data.
Gelb adds that "It is clear that when a ruler writes of having built a temple for a certain
divinity, he means not only that he erected a temple, but also that he provided it with all the
necessary means of social and economic support" ( 1 979b:3).
Lesson 1 29
Text 1
~
trr ~
m w P>
*" re:
5
. r=Cl
rmEf
---
W P --
t>
30 Lesson 1
Notes: autographs
When obtainable, photographs of the texts used in the Lessons have been included.
This has not always been possible or desirable, and so most of the texts are presented as
"autographs". In Assyriological parlance, autograph refers to the hand-copy done by a
modern Assyriologist, to imitate the cuneiform. The quality of autographs can range from
very accurate to very poor. To quote Lieberman,
It is, of course, patent that the "autographs" of all copyists are not equally
reliable. Their objectives, ranging from an exact reproduction including
every scratch on the tablet to a highly abstract conventional representation of
the original (some Assyriologists are even known to have produced "copies"
from their transliterated notes) as well as their individual skills and abilities
make the value of their copies diverge (1977: 67).
It is only through long experience that one gets a feel for how accurate certain
Assyriologists are (or aren't) in their autographs.
-Writing practices
Both Sumerian and Akadian are written from left to right across the writing surface.
(The earliest Sumerian texts were inscribed in vertical columns, read from right to left.)
Most royal inscriptions are subdivided into "lines", marked by an actual line drawn or
impressed on the writing surface. The use of such lines in Sumerian (and Akkadian) is to
some extent dependent on the genre of text; royal inscriptions, for example, use them
regularly. Many literary texts use them, but just as many do not.
There is some oscillation in the use of the word line. This particular text was divided
by its scribe into seven units, but the fifth of these units actually contains two rows of text.
In order to be precise, some Sumerologists use the term "case" or "register" to describe the
units physically demarcated by the scribe, and the term "line" to describe the actual rows of
signs. Thus, in this text case 5 has two lines. Although this is a very handy distinction,
most scholars, will, in fact, simply use the term line to mean both line or case, especially in
k
unambiguous contexts; this is the rocedure followed here.
In line 5, the determinative 1 begins the second line within the case. There are six
cuneiform signs in this particular expression. It would have been physically impossible to
put all these six signs on one line, so the scribe put them on two lines. If he had put the ki
sign with the Urims -sign, there would have been too much empty space on the second line
of the case. By indenting the second line of the case, the signs representing the ON are
grouped in close proximity to each other.
genitive marker is lalc! after a consonant, Ikl after a vowel. Ikl is one of the amissable
consonants, and hence does not appear in writing in word-final position. The lal of the
genitive marker is usually contained within a sign which reduplicates the consonant
immediatelypreceding the/a/.
3. nin-i!-ni = nin.ani.(r). S umerian has a set of suffixes to indicate pronominal possession.
They are referred to as "possessive-suffixes" or "pronominal suffixes". .ani is the
possessive-suffix marking third person singular. The fonns of the first and second per
sons, and of all the plurals, are discussed later.
S ince Sumerian has no gender system, .ani can mean either "his" or "her". However,
Sumerian does have remnants of what is usually referred to as a distinction in "animacy".
Human beings are "animate"; things and animals are "inanimate". In the case of the
possessive-suffix, .ani is only used to refer to animate antecedents; an entirely different
fonn (.bi) is used to refer to inanimate antecedents (corresponding to English "its").
After a consonant, the suffix appears as: .ani. After a vowel, it appears both as: .ani,
and as: .ni. For example, "his house" can appear as both -i!-ni and -ni; in the Ur III royal
inscriptions, the fuller spelling is much more common .
.ra is the case-marker for the dative case. Its fonn is Iral following a consonant and
Irl following a vowel.
Case-endings in Sumerian work differently than they do in the Semitic or the Indo
European languages. In Sumerian, case-endings occur at the end of an entire nominal
phrase. A nominal phrase can vary in size. Minimally, it can consist of a single noun. It
can also consist of a noun with a possessive-suffix, or with an adjective, or with an
embedded genitive phrase, or even with a long series of appositives. In this particular case,
the nominal phrase spans lines 1 to 3. It consists of: a divine name (Nanna); an appositive,
consisting of a genitive phrase (nin.an.a(k ; a second appositive, consisting of a noun with
a possessive-suffix (nin.ani). The dative case-marker .r comes at the end of this entire
phrase. This can be diagrammed as: [Nanna nin.an.a(k) nin.ani].r. This is, in general, the
way all case-markers work in S umerian (and, even more generally, in agglutinative
languages).
The dative case is primarily used in S umerian to express an indirect object; for
example, "He gave it to the king". It is also frequently used (as it is here) to express a
benefactive, that is, the person on whose behalf an action was perfonned. In such cases, it
can be translated by "for".
The case-marker Irl is not written here. Its presence in spoken S umerian is shown by
the fact that it is actually written in other (mostly later) inscriptions. In these other
inscriptions, there are fonns such as: digir-ra-ni-ir, "for his god" = digir.ani.r (following the
nonnal convention that CV-VC stands for ICVC/, that is, ni-ir = Inir/). In the body of
texts in this book, .r first appears in Text 14, an inscription of Amar-Sin, the grandson of
Ur-Nammu.
It is not known why the Irl is not written; this is discussed in Lesson 14. The
situation is different from that of the genitive marker. The Ikl of the genitive marker is an
amissable consonant, and so is regularly not written. But Irl is apparently a non-amissable
consonant, and does occasionally appear in the writing.
Lesson 1 33
4. ur followed by the name of a deity is a very common way to form personal names in
Sumerian, in all periods of the language. In Text 1 , Ur_dNammu occurs; in Text 1 9a, Ur
dLamar. Such names are genitive phrases, meaning "man of DN" or "warrior of DN".
The name then is to be understood as: Ur.Nammu.(k), with the genitive marker taking the
form Ikl after a preceding vowel. However, there is some disagreement about the presence
of the genitive marker in proper names. Some Sumerologists believe that in proper names
the genitive marker was deleted. Thus, this particular name may have been pronounced as
lumammu/, and not as lumammuk/. Other Sumerologists however, do not believe this to
be so. The first practice has been followed in this book, and thus this name has been
transcribed as Umammu, not as Ur.nammu.(k). In translation, the most common
Assyriological practice is to give the name as Ur-Nammu.
5. l.Y.gal-Urims ki-ma-ke4 = lugal.Urim.ak.e, "king of Ur". Because the genitive marker
follows a consonant (here, Im/), its full form (/ak/, with initial la/) is used.
When the genitive marker is directly followed by a vowel, the Ikl is pronounced, and
shows up in the writing (recall that in such phrases as nin.an.a(k), the Ikl is word-final,
and hence does not show up in the writing).
The .e is the marker of the ergative case, as discussed under Ergativity. As do all
case-markers, it comes at the end of the entire nominal phrase. The nominal phrase here
consists of a personal name, Ur-Nammu (a genitive phrase in origin), and an appositive
consisting of a genitive phrase, lugal.Urim.ak. This may be diagrammed as: [Umammu
lugal.Urim.ak].e.
The ergative case-marker .e marks what we would call the active subject of a transitive
verb, or, in more appropriate terminology, the agent. (Because of inconsistencies in
terminology, however, this .e is sometimes referred to as "agent", "agentive marker" or
"ending", "subject", "transitive subject", "ergative marker", etc.)
The cuneiform signs do not reflect well the morphology of Sumerian here. In
transliteration, the signs are: lugal-Urimski-ma-ke4' In morphological transcription, this is:
lugal.Urim. ak.e. The rna-sign reduplicates the final Iml of Urims, and includes the lal of
the genitive marker. The ke4 -sign includes the Ikl of the genitive marker, and the I el of
the ergative case-marker. Thus, both the rna-sign and the ke4 -sign represent segments of
two different morphemes. This use of the ke4 -sign is very frequent; it is the sign normally
used for the combination of segments of the genitive marker and the ergative case-marker.
Not much is known about the syllabic structure of spoken Sumerian, but it may have
been closer to the written form than to the morphological transcription. This line may have
been syllabified something like: Ilu-ga-Iu-ri-ma-ke/. If so, the written form is actually
closer to the presumed syllabic structure of Sumerian than it is to the morphemic structure
of Sumerian.
The use of hyphens in transliteration varies to some degree from scholar to scholar.
All S umerologists would use hyphens in the wo.rd Urims ki-ma-ke4' Some would put a
l
hyphen between lugal and Urims : lugill-Urim l-ma-ke4' In this latter case, hyphens are
being used to link all the signs which form the entire nominal phrase. Others use hyphens
only between signs belonging to one word. It is not always easy, however, to define
"word" in Sumerian.
34 Lesson 1
6. --ni = e.ani.0, "his temple". As in line 3, .ani is the third person animate possessive
suffix. The antecedent is ambiguous; it could refer to Ur-Nammu, or it could refer to
Nanna. From other texts it is clear that .ani refers back to Nanna.
The .0 is the case-marker for the absolute case. This case indicates what we would
call the direct object of a transitive verb, or, more appropriately, the patient. There is,
however, not a great deal of consistency in nomenclature, and so such terms as
"accusative", "direct object marker", etc., are commonly used.
The nominal phrase here is quite short, consisting of the noun , and the possessive
suffix .ani: [e.ani] .0.
7. mu-na-du = mu.na.(n.)du.0, "he built". This line contains the verbal phrase. The verb
in Sumerian works rather differently than the verb in the Semitic or Indo-European
languages. A finite verb form in Sumerian consists of a series of verbal prefixes, followed
by a verbal root, then followed by a smaller series of verbal suffixes. Certain of these
affixes are obligatory, while others are optional. Because of the general uncertainty of
Sumerian grammar, the precise number of prefixes occurring before the verbal root is
unsure. The view presented here might be called "minimalist". Alternative interpretations
will be discussed later.
The entire sequence of verbal prefixes occurring before the verbal root is usually
referred to as the "verbal chain". The first prefix to appear in this chain is an optional
"modal-prefix" (also referred to as a "mood-marker"). Modal-prefixes are used for such
sentence types as cohortative, jussive, subjunctive, etc. A "normal" declarative sentence is
in the indicative mood, which is unmarked. The verb in line 7 is indicative, and so there is
no modal-prefix.
The second position is occupied by the "conjugation-prefix". There are some half
dozen conjugation-prefixes. These prefixes are among the most mysterious features of
Sumerian; it is not known exactly what information these prefixes convey. This means
that it is not known, for example, what the difference in meaning is between a finite verbal
form with the conjugation-prefix mu and one with the conjugation-prefix i. Such variation
occurs in the texts, but it is not known what this variation implies.
Needless to say, there are several theories about the function of the conjugation
prefixes. They may be connected with time: indicating whether events are near or far
(temporally, or even emotionally) relative to the speaker. They may have to do with space:
indicating whether events are near or far (spatially, or even emotionally) relative to the
speaker. At times, they seem to correspond to a polite - familiar distinction.
It is probable that the conjugation-prefixes convey nuances which are not normally
conveyed in English. This means that even if it were understood what the conjugation
prefixes meant, it would not be possible to translate them readily into English, except by an
elaborate periphrasis. (Jacobsen, for example, believes that the conjugation-prefix mu is
used "To indicate 'closeness' to the speaker if by closeness we understand not only
closeness in space and time but also emotional closeness, empathy, involvement"
[ 1 965: 4 37].)
In practice, Sumerologists ignore the conjugation-prefixes; they are not reflected in
translation. Writing in 1972, Maurice Lambert said: "Today, the prefix does not exist for
Lesson 1 35
the translator of S umerian, it is only an object of study for the grammarian" ( 1 972-3:97).
The problem of the conjugation-prefixes cannot be solved here. In subsequent texts,
the various conjugation-prefixes will be pointed out, and the possible kinds of information
which they may be conveying will be discussed.
Text 1 uses the conjugation-prefix mu. This conjugation-prefix is very common in the
Ur III royal inscriptions. In fact, almost all past-tense verbs in main sentences in the Ur ITI
royal inscriptions use the conjugation-prefix mu.
The next set of prefixes are the (mostly) obligatory "dimensional-prefixes". There is
nothing comparable to these forms in Semitic or Indo-European. They "cross-reference"
(or "resume" or "register") the case relationships appearing in the various nominal phrases
in the sentence, with the exception of the agent and patient. In the verb in line 7, the
dimensional-prefix .na cross-references the dative case marked by .r in line 3.
Most earlier studies of S umerian stated that the dimensional-prefixes were obligatory,
and that there was a one-to-one relationship between case relationships and dimensional
prefixes: every case relationship is resumed by its dimensional-prefix, and conversely
every occurrence of a dimensional-prefix implies a corresponding case relationship
somewhere in the sentence. While this one-to-one correspondence may have been valid for
"pre-historic" Sumerian, in actual historic Sumerian the situation is not so neat. Gene
Gragg has made a detailed study of the dimensional-prefixes in the Old Babylonian literary
texts; he states that they "function independently of concord to a much greater extent than
has been recognized by current theories" ( 1973a: 10).
The dimensional-prefixes often seem unnecessary or redundant, because they do not
convey any new information; rather, they "merely" cross-reference the already-present case
relationships. However, all languages have a certain amount of built-in redundancy, to help
cope with the possiblities of information being garbled or lost. Many other languages
cross-reference case relationships, in various ways.
The nominal phrase in the dative is the only nominal phrase (except those indicating
the agent and the patient) in the sentence, so only one dimensional-prefix occurs. If other
nominal phrases were present, they would also be resumed. Thus, it is possible for there to
be one, two, or three dimensional-prefixes in one verbal chain; that is, the dimensional
prefixes are cumulative. (The longest attested sequence appears to be four dimensional
prefixes in one verbal chain.) There is a hierarchical order to these prefixes; the dative, for
instance, always comes first. Not all such rules, however, are understood; in addition, there
are certain morphophonemic changes which are not clear. These complications will be
discussed later.
Following the dimensional-prefixes comes a (probably) obligatory prefix, the
"personal-affix" (there is no generally-accepted term). These forms have been much
discussed. They apparently cross-reference the agent and the patient, although this is not
completely certain.
In the case of a verb in the past tense, the personal-affix in this position cross
references the agent. Thus, in Text 1 , the personal-affix .(n) cross-references the agent
marked by the ergative case-marker of line 5.
The form of the third-person singular animate personal-affix is: .n. As will be seen
36 Lesson 1
later, the personal-affix has different forms for first and second person, and also different
forms for inanimate agents.
This particular prefix always occupies the position closest to the verbal root.
However, this prefix frequently does not show up in the writing. The reason for its
absence is not as clear as that of, for example, the dative case-marker. The dative case
marker is not normally written in texts from the early stages of the Ur III dynasty (nor in
earlier texts), but it begins to show up frequently in texts from the time of Amar-Sin on.
Thus, scholars are reasonably confident that the Irl of the dative case-marker is present,
even if not written; its later appearance is the result of a change in orthographic practice.
The rules governing the presence and absence of the personal-affix 1nl are, however, not
so clear-cut; it is not often written even in later texts.
The presence or absence of .n cannot simply be correlated with a dimension of time.
In the Gudea texts, for example, forms both with and without .n occur, with no obvious
rules governing their distribution. And in later S umerian, forms also occur both with and
without the .n. This means that rules cannot yet be determined for the presence or absence
of 1nl in the script, and it is not in fact sure at what level such rules would apply. The rules
may be purely orthographic; there seem to be other cases in Sumerian where syllable-final
nasals are not expressed in writing. Or, the rules may be phonological; the 1nl may have
dropped early, leaving a nasalized vowel, which could not adequately be represented in the
script. More probably, there may be a complex set of morphological and syntactical rules
governing deletion of 1nl; it has been posited, for example, that 1nl is only used (and so
only expressed in writing) to resolve possibly ambiguous cases.
Partially for convenience sake, I have assumed that the personal-affix .n is always
present, unless there is a specific reason for its absence. Hence, it is transcribed as: (n.).
This presumed consistency must be taken with a grain of salt.
After all these obligatory and optional prefixes, comes the verbal root, du in this
particular case. The root in Sumerian appears to be invariable. There is nothing like the
complicated inflection of Semitic or Indo-European roots for person and number (the only
inflection for person is in the personal-affix position, immediately before the verbal root; a
limited inflection for number occurs in a set of personal-affixes after the verbal root).
There is no canonical shape of the root. Roots of the syllabic shape ev and eve are
perhaps the most common, but roots of other syllabic structures are frequent.
After the verbal root, there occur a number of optional affixes, not all of which are
well-understood. Some of these affixes are used to express modal and other nuances, such
as potentiality, irrealis, etc.
For a verb in the past tense, the most important affix which occurs in this position is
the personal-affix which cross-references the patient. The personal-affix which cross
references a third-person singular patient can be represented by zero, .0. Thus, the patient
in this sentence (e.ani.0) is resumed by a .0 after the verbal root. This means that the
patient is marked by .0, and that it is cross-referenced by .0. This may vaguely seem like
cheating ("nothing resumed by nothing"), but there are theoretical justifications for this
interpretation.
Thus, the agent and the patient are resumed differently: The agent is resumed in the
Lesson 1 37
position immediately before the verbal root, and the patient is resumed in the position
immediately after the verbal root. The term personal-affix is used to refer to both affixes.
To sum up, the verbal phrase in Sumerian normally consists of: an optional modal
prefix (the indicative is unmarked); an obligatory conjugation-prefix, whose function is
unclear; one or more basically obligatory dimensional-prefixes, which cross-reference all
case relationships (except that of the agent and patient); an obligatory personal-affix, which
in the past tense cross-references the agent; the verbal root; an obligatory personal-affix,
which in the past tense cross-references the patient; other optional affixes.
This particular verbal form may be summarized as follows:
mu na . (n.) du . 0
(1) (2) (3) (4) (5)
(1) conjugation-prefix
(2) dimensional-prefix cross-referencing the dative
(3) personal-affix cross-referencing the agent
( 4) verbal root
(5) personal-affix cross-referencing the patient.
The verb in line 7 was translated as past tense, without any discussion. Sumerian has
two sets of verbal forms. The difference in function between the two is somewhat unsure.
Some Sumerologists believe that the difference was one of tense (past - present-future);
others believe that it was a difference of aspect (perfect - imperfect); and others believe that
it was a difference of Aktionsart (punctual - durative, etc.). For convenience sake, they will
be referred to here as aspects.
Akkadian scribes gave names to these two aspects. One aspect they called bamm
("quick"), and the other they called maru ("fat"). There is some evidence that the S umerian
word for bamm was Y4 ' and the word for marii was niga; the original meaning of these
two words is not sure. The terms bamn! and marii are frequently used by modem Assy
riologists when referring to these verbal forms in Sumerian.
In the Ur III royal inscriptions, it does seem that basically the bamt.!! is used for
actions which occurred in the past, and the marii is used for actions in the present and
future. That is, the two seem more tense-like than aspect-like. But this may be due to the
relative simplicity of these inscriptions.
The bamn! form is unmarked; it is the citation form (the form given, for example, in
the Vocabularies). As will be seen later, the marii is formed from the bamt!! in several
different ways, and the systems for cross-referencing the agent and patient in the bamt!! and
in the marii are quite different.
Discussion: structure
Having examined this inscription with a fine-tooth comb, let us now consider the
structure of the inscription as a whole. If all appositional noun phrases are grouped with
their head nouns, and their functions are labeled, we see:
38 Lesson 1
The dative marked in .r is resumed by the dimensional-prefix .na; the ergative marked
in .e is resumed by the personal-affix .n; the absolute marked in .0 is resumed by the
personal-affix .0. This is a rather aesthetically satisfying system; as will be seen later,
however, things often do not hang together so neatly.
Second, let us look at the word order:
This particular order is actually somewhat different from standard Sumerian syntax.
In more standard Sumerian, the word order is:
The difference in word order between standard Sumerian prose and that of the royal
inscriptions is in the position of the benefactive. In royal inscriptions, the benefactive is
almost always fronted; this gives added emphasis to the deity on whose behalf some act is
being commemorated. In English, the difference might be reflected as "For Nanna, Ur
Nammu built his temple", instead of "Ur-Nammu built his temple for Nanna".
Hallo's investigation of the structure of the Ur III royal inscriptions showed that their
style is very formulaic. A typical inscription is composed of the following elements, almost
al ways in the same order:
(1) A benefactive phrase, giving the name of the deity, with optional epithets;
(2) An agentive phrase, giving the name of the builder or donor, with optional
epithets;
(3) A patient phrase, describing the object built or donated;
(4) A verbal phrase, highly stylized and formulaic.
-Terminology
As does any discipline, Sumerology has engendered its own host of technical terms,
such as Auslaut, amissable, etc. Some of these terms are peculiar to Sumerologists; they
are not standard terms familiar to general linguists. Unfortunately, some of these terms are
used in ways which cause general linguists to take umbrage.
The term verbal chain is used here to refer to the series of prefixes which occur before
the verbal root. Other people use the term to include the entire verb: prefixes-root-suffixes.
Lesson 1 39
Similarly, the tenn nominal chain is sometimes used to refer to a nominal phrase.
Sometimes both the verbal chain and the nominal chain are subsumed under the category
"Kettenbildung".
The term conjugation-prefix, in particular, is misleading, because these elements have
nothing to do with conjugation, as this tenn is usually understood. However, this is the
only term used by Sumerologists.
There is no standard tenn to refer to what is called here the dimensional-prefix; the
most common tenn is probably dimensional infix. This use of the tenn infix, however, is
often irksome to general linguists, who use the tenn to refer explicitly to an affix placed
within another morpheme; an example would be the It/ in the Akkadian Bt stem, or the
Arabic Eighth Form.
-ke4
The value of theiMJ sign as ke4 was deduced by Kramer in 1936. There is no native
grammatical tradition which gives this value; the Akkadian lexical lists give the values of
this sign as g and k4 (in addition to such values as kid, etc.) Kramer reasoned that the
only way to make the Sumerian writing be consistent with our understanding of the
morphology of the Sumerian genitive was to posit a reading ke4 ' even if the lexical lists do
not give this value. Virtually all modern scholars have accepted his reasoning. (However,
even this seemingly well-established fact of Sumerian grammar has recently been
questioned, by Lieberman. He believes that the genitive marker was /ag/, not /ak/, but he
has not yet published his reasons for doubting the conventional interpretation.)
-Animacy
As was mentioned when discussing the possessive-suffix .ani, Sumerian has traces of
an animate - inanimate distinction. This distinction is also seen in the personal-affix of the
third-person bamt!!-transitive verb, where .n marks an animate agent, but .b marks an
inanimate agent (rather a rare occurrence). This animate - inanimate distinction does not
carry through all aspects of the grammar.
The tenns animate and inanimate are those traditionally used by linguists, even if this
means that animals are called inanimate (Jacobsen prefers the tenns "personal" and "non
personal"). In fables, however, animals are usually treated grammatically as animate.
-Conjugation-prefixes
Lambert was quoted above, to the effect that the conjugation-prefixes are simply not
tran slated. This is because it is not known what infonnation they convey, and the odds are
that their function has no easy equivalent in English. Edmond Sollberger has said:
Their true rOle is so distinctively Sumerian, they express ideas so alien to our
languages, that not only is there no consensus on the nature of their function,
but we simply ignore them without impairing, or so it seems to us, our
understanding of the text. There is no other translation for mu-gar and i-gar
than "(he) placed", although it must be pretty obvious that had there been no
difference there wouldn't have been two prefixes . ... It is legitimate to posit
40 Lesson 1
that a certain verbal fonn implies that the action is perfonned by the subject
wishing to indicate that his goal, though within his immediate perception,
remains without his actual sphere of physical contact; it is another thing to try
and express that in one good English (or even Gennan) word (1973:160-61).
F.R. Kraus has criticized this view of Sollberger: "Sollberger's opinion, that Sumerian
texts can be understood without paying attention to the verbal prefix, is valid for a certain
kind of text, but is certainly not valid for legal documents" (1958:83 n.47).
- Conjugation
The fonns of the bamm-transitive verb in the singular are listed here. This and other
paradigms should be understood as reflecting Ur III morphology, in Ur ITI orthography.
The model verb used is sar, "to write", with the conjugation-prefix mu.
The fonn of the first person is somewhat unsure. The fonn of the second person is
more sure, because the .e sometimes shows up in the script. Similarly, the fonns of the
third person are "sure", because of the occasional presence of .n and .b in the script.
In this section, the personal-affixes .n and .b have been discussed as markers for the
third person. Earlier, it was said that they cross-reference the agent. S trictly speaking, they
cross-reference a third person agent. A first person agent ("I") is cross-referenced by . 0,
and a second person agent ("you") is cross-referenced by .e. In other words, one can
understand the personal-affixes as cross-referencing the agent, or as marking the person of
the verb; in Sumerian, these are two different ways of describing the same thing.
The first person to recognize that bamm and marii were used as native grammatical
tenns was Heinrich Zimmern, in 1 885, although he did not know what they meant. Paul
Haupt was apparently the first to give these words their etymologies as "quick" and "fat",
in 1932.
- Typology
- Function of text
Let us now look at the function and Sitz im Leben of this particular text. Hallo has
divided the Ur III royal inscriptions into five categories, based on typological criteria:
standard, building, votive, weight, and seal inscriptions. Text 1 is a building inscription;
examples will occur of all the other four types. Building inscriptions are defined by Hallo
as "monuments that became integral parts, whether functional or decorative, of the
buildings which they commemorated" (1962:8).
The building inscriptions are further subdivided on the basis of the type of object they
were inscribed on: bricks (the most numerous of all royal inscriptions), foundation
deposits, door sockets, and clay cones. Examples will be seen of each. Text 1 was
inscribed on a brick, forming an actual part of the masonry of a building. Building
inscriptions in general were not designed to be read by the builder's contemporaries; rather,
they were designed to be read by future rebuilders of the building, most likely future kings.
Ultimately, these buildings and their accompanying inscriptions can be thought of as
attempts by rulers to attain some form of immortality. (Text 16 is a door socket with two
inscriptions. One is of an early ruler of Ur [about 2400 BC]; the other is of a ruler of the
Ur ill period. The door socket was evidently uncovered during rebuilding carried out in
the Ur III period, and was re-used.) As will be discussed in Lesson 2, often several copies
of the same inscription are found.
- History
photo is of the remains now standing; the condition of these remains is partially a result of
modem reconstruction of the site.
. "
uf ':.:
" '
" --
----
-- --
-- '--
----
--'--
--'
. . .
Lesson 1 43
While the sacred area as a whole was dedicated to Nanna, he also had his own coun in
front of the ziggurat, and other buildings sacred to him. The entire sacred complex was
known as the E-ki-nu-Ml; the ziggurat was known as the E-temen-ni-guru3 (see Lesson
9). Both tenns are of uncertain etymology. The brick containing Text 1 fonned pan of a
temple known as the -bur-sag ("mountain temple").
One of the more famous pieces of Ancient Near Eastern art is known as the "Stela of
Ur-Nammu". It was found in a very fragmentary state in Ur, scattered throughout the
Nanna temple complex; it may have been destroyed during the Elamite sack of Ur in 2004
BC. It depicts a number of symbolic activities, mostly obscure to us, but apparently shows
Ur-Nammu himself carrying building tools (his name appears on a floating fragment of the
stela). This stela has been known since the 1920s, but restoration work is still on-going.
(A very interesting discussion is in Canby 1987.)
44 Lesson 1
-Literature
- Proper names
Most recently, Miguel Civil ( 1985:27) transliterates the name of the founder of the Ur
III Dynasty as Ur-Namma, instead of the usual Ur-Nammu. He bases himself on
attestations of the name in syllabic orthography. He suggests that the original form of the
name was a theoretical IUr-Namnam/. Jacobsen also now reads the original form of the
divine name as Namma, but derives INammal from Inin inim/, "lady female genital";
INammu/ is a later form ( 1987: 155 n.5).
As will be discussed in Appendix 2, a number of bilingual lexical lists have been
found at Ebla. Names of gods occur several times in these lists. For Nammu, the Eblaite
equivalent is given as: i-nu bf!.-mi-um. This is somewhat difficult to understand. F.M.
Fales thinks that the Eblaite expression might mean "venemous tooth" ( 1 984: 176). It is
hard to square such a description or epithet with what is known about Nammu.
-Titulature
Many of the appositive phrases describing the king in these inscriptions are actually
titles, occurring in many inscriptions. (Although sometimes it is not possible to tell if an
adjectival phrase is a title or not.) Much work has been done in determining the origin of
certain titles, their relationship to parallel Akkadian titles, their falling into desuetude, etc.
Lesson 1 45
The principal work on this topic is by Hallo: Early Mesopotamian Royal Titles: A Philo
mic and Historical Analysis (1957). This was followed by M.-J. Seux in 1 967, who
studied in particular the individual words occuring in Sumerian and Akkadian titles.
The title used in Text 1 , "King of Ur", was used by all five kings of the Ur III
Dynasty.
Lesson 2
47
48 Lesson 2
Notes
Inanna The daughter of Nanna. She was the Sumerian goddess of love and fertility, of
the morning and evening star, and to some degree of war; she had other sides as well. She
may have absorbed some of the attributes of originally independent deities. Later equated
with the Akkadian Ishtar, in some ways she was the most important goddess in the
Mesopotamian pantheon. Because of her rather fiery temperament, and the manifold
aspects of her personality, she is perhaps the most interesting of all Mesopotamian deities.
She was worshipped in many cities, but especially in Uruk, where she was the tutelary
goddess. Her principal temple at Uruk was the Eanna -an-na = e.an.a(k), "house of the
sky/heaven").
The reading of her name is much disputed. It is also transliterated as: Inana, Innin,
and Ninni6' The original pictographic meaning of the cuneiform sign is also uncertain. Her
name is usually interpreted as: nin.an.a(k), "Lady of the sky/heaven". This is also how the
Akkadian scribes understood her name. Jacobsen believes that Inanna was originally the
"numen of the communal storehouse for dates". He thinks that the / ani-component of her
name meant "date-clusters": "Her name ... would appear to have meant originally 'the lady
of the date-clusters'" ( 1957 : 1 08); later, her name was "re-interpreted" as "lady of the
sky/heaven".
Ki-en-gi This ON is always written syllabically. The etymology is unsure; this is dis
cussed below. The word ended in a /r/, not reflected in the script. The Akkadian equi
valent of Kiengi was Sumeru. This Akkadian word may be a dialectal pronunciation of the
word Kiengi(r). The English word "Sumer" is usually thought to derive from the Ak
kadian form.
The first appearance of Ki-en-gi is in an inscription of Enshakushanna of Uruk (who
ruled approximately 2432-2403 BC), who refers to himself as: en-Ki-en-gi l!!gal-kalam
ma, "the lord of Sumer, the king of the land".
Ki-uri The etymology is unknown. It is not impossible that the ki-element was originally
a determinative.
nitab The basic meaning appears to be "male"; it can often be loosely translated as "man".
The Akkadian equivalent is zikaru, glossed by the CAD as: " 1 . male (human and animal),
2. man, 3. ram".
kalag The Akkadian equivalent verb, dananu, is translated by the CAD as: "to become
strong". The verbal adjective, dannu, is translated as: " 1 ) solid, strong, hard, heavy, thick,
massive, fortified, steady, loud, 2) legitimate, binding, reliable, 3) strong, powerful, mighty,
great, 4) fierce, savage, difficult, dangerous, serious, grave, obstinate, bad, tyrannical,
harsh, pressing, urgent, essential, imperative".
Lesson 2
Text 2
* tH bJHf>
1=1
rt Jf> r-i
1\ Bm
l!:f *
mI3 ff J?>
IB *7
Notes
Some of the signs which occur both in Text 1 and in Text 2 differ slightly from ea
other. In Text 2, the -sign and the ke4-sign differ only in their length. In Text 1 they we
of the same length, but differed in the position and length of the verticals. Strictly speakir
the sign-shapes in Text 2 are more "correct".
In line 6, the word Ki-en-gi is divided into two lines within the one case.
50 Lesson 2
sign would be a kind of "phonetic complement": It gives some extra information to the
reader, helping him to choose the correct reading of the previous sign. The word might
then be transliterated as: kalagga.
A second view attempts to make the signs approach the transcription. Since this word
is pronounced Ikalaga/, and since the Igal is expressed by the g-sign, this view says that
the first sign must therefore be read Ikala/: kala-g. Thus, this view really derives the
transliteration from the transcription.
The third view says that the transliteration should not necessarily be expected to fit the
transcription. Rather, there are certain general rules of Sumerian orthography which are
found in several different contexts. In this particular case, the transliteration kalag-g!!.
reflects the orthographic rule that a consonant is graphically reduplicated before a word
final (occasionally syllable-final) vowel, particularly across a morpheme boundary. For
example, in Text 1 there occurred: nin-an-na, for nin.an.a(k).
The entire problem is not easy to resolve. Several obvious questions come to mind:
How can one know, for instance, that the sign ID can, in fact, be read as kal, or kala, or
kalag, or kalag, or kalg? To what extent are readings "manufactured", to make the trans
literation more closely approximate the transcription? How valid is the general rule of
Sumerian orthography presented above?
In practice, inconsistencies arise in transliteration, because no matter which
transliteration system is followed, the meaning is normally clear. Whether these two signs
are understood as kalagga, kalag-g or as kala-g, everyone would understand the pronun
ciation to be Ikalaga/, and the meaning to be "mighty". (Even here, however, some S ume
rologists would say that the original form */kalagal > Ikalga/. It is true that similar cases
of vocalic loss are attested in Sumerian. However, the [late] syllabic writings of the type
kal-Ia-g!!. would seem to argue against such an interpretation in this particular case.)
Therefore, some Sumerologists prefer not to deal with these problems, unless they are
interested in the writing system per se.
This problem has been discussed at some length, because it is useful to be aware of the
theoretical principles which underpin our understanding of the writing system. This type of
knowledge is also essential if one is to understand borrowings of the Sumerian writing
system, such as, e.g., the writing system used for Eblaite. And, it is important to be
prepared for (and to understand the reasons for) the inconsistencies and variations in
transliteration which are encountered in Sumerological literature.
In general, adjectives in S umerian follow the noun they modify.
5. illgal-Urims ki-ma = lugal.Urim.a(k), "king of Ur". Just as both kalag- and kala-g
are found in transliterations of the same two signs, so also these signs are found trans
literated as Urimski-ma and UriskC ma.
6. lugal-Ki-en-gi-Ki-Uri-ke4 = lugal.KiengLKiuri.k.e., "king of Sumer and Akkad".
Sumerian has a conjunction meaning "and", linking nouns, but it is relatively
uncommon. Instead, Sumerian prefers to conjoin two nouns directly: an-ki "heaven and
earth".
The first element of the genitive phrase is the single noun lugal. The second element is
formed by the two conjoined nouns, KiengLKiuri. The genitive marker .k follows the two
52 Lesson 2
elements. This can be diagrammed as: lugal.[Kiengi.Kiuri] .k. It is possible for either
element of a genitive phrase to be even more complex, consisting of a noun with a
possessive-suffix, an adjective, a relative clause, etc.
Lines 3-6 form a long nominal phrase, ending in the ergative case-marker .e. This
nominal phrase consists of: a personal name (line 3); an appositive consisting of a noun
and an adjective (line 4); a second appositive consisting of a genitive phrase (line 5); and a
third appositive, consisting of a more complex genitive phrase (line 6).
Discussion: structure
It is instructive to compare the structure of Text 1 and Text 2:
Text 1 :
[Nanna, nin.an.a(k), nin.ani] .(r) benefactive
[Urnammu, lugal.Urim.ak] .e agent
[e.ani] .0 patient
mu.na.(n.)du.0 verb
Text 2:
[Inanna, nin.ani] .(r) benefactive
[Urnammu, nitag.kalaga,lugal.Urim.a(k), agent
Iugal.KiengLKi uri.k].e
[e.ani] .0 patient
mu.na.(n.)du.0 verb
The order of the constituents is the same. As mentioned in Lesson 1 , the constituent
order in these inscriptions is quite formulaic. The difference in the two inscriptions is in the
length of the various nominal phrases, and not in the basic structure.
- Brick-stamps
The cuneiform signs in this text are much more "linear" than those of Text 1 . This is
because Text 1 was "handwritten" by a particular scribe. Text 2 was produced by a "brick
stamp". Brick-stamps were used to mass-produce copies of inscriptions. The writing on
them is done in reverse ("mirror-writing"), so that the impression comes out correctly. The
shape of the signs used tends to be linear, although occasionally they can approach the
shape of the handwritten signs. Several brick-stamps have been preserved, although
apparently none from the Ur III period. The following illustrations are of brick-stamps
from the Old Akkadian period:
Lesson 2 53
-Case relationships
.ra and the other case-markers in Sumerian are variously referred to as "cases", "case
markers", "case-endings", "postpositions", "postfixes", etc. Strictly speaking, these tenns
are not all synonymous, because they do not all refer to the same level of analysis.
The tenn "dative case", for example, refers purely to a grammatical relationship. This
case can be used to indicate several different semantic relationships: indirect object,
benefactive, etc. "Dative case-marker" or "case-ending" refers to the specific fonnal device
which signals this grammatical relationship, that is, the .ra. "Postposition" or "postfix"
means that the case-marker occurs at the end of a nominal phrase. (This contrasts with
English, for example, where "prepositions" occur in front of a nominal phrase.) Thus, in
Text 1 and 2, .ra can be described as a postpositive case-marker of the dative case, used to
express the bepefactive.
Although these tenns are distinct, in practice they are often used somewhat
indiscriminately. This is because it will normally be clear from the context which level of
analysis is being referred to. Similarly, the dimensional-prefixes are sometimes said to
54 Lesson 2
cross-reference the cases, and at other times are said to cross-reference the case-endings.
Strictly speaking, they cross-reference the case relationships which are marked by the case
endings. For ease of exposition, however, it is usually easier to present them as cross
referencing the case-endings themselves.
The genitive does not behave like the (other) cases in Sumerian, and so it is
occasionally referred to as a "genitive marker", instead of as a case. First, a genitive phrase
can be embedded within a nominal phrase, which can then have its own case-marker. That
is, the genitive can be cumulative with respect to the (other) cases. For example, the
genitive can be directly followed by the ergative case-marker .e, as in Text 1 and Text 2.
The (other) cases, however, are not cumulative with respect to each other. If a nominal
phrase has the dative case-marker, for example, it is impossible for it to have any other
case-marker. Second, the genitive is not resumed by any dimensional-prefix. The dative,
for example, is resumed by the dimensional-prefix .na. However, the genitive is not
resumed.
The reason for the difference in behavior is because of the different r6le which the
genitive plays in a sentence. Genitives relate noun phrases to noun phrases. But the (other)
cases relate noun phrases to verb phrases. That is, genitives and cases perform two
different functions. However, "case" is the term most frequently encountered in
Sumerological literature.
Some scholars use the term "adnominal" case to refer to the genitive and to the
equitative (to be mentioned later). Both can be cumulative, and neither is resumed by any
dimensional-prefix.
The ergative and absolute cases pattern together, in that they are the only cases cross
referenced in the immediately pre- and post-verbal root slot. (In some ergative languages,
verbal cross-referencing only occurs with the agent and the patient, and not with any other
case relationship.)
The cases besides the ergative, absolute, genitive, and equitative are referred to as
"adverbial". They include the following; they will be studied in subsequent lessons:
dative; terminative; locative; locative-terminative; comitative; ablative. Some scholars use
the term "oblique" instead of "adverbial"; others use the term "dimensional". The latter is
rather nice, since these cases are the only ones to be cross-referenced by the dimensional
prefixes.
To sum up, the Sumerian cases may be categorized as:
It was Poebel who definitively established the form and function of the S umerian
genitive ( 1 935). Earlier views were quite different. For example, Franois Thureau-
Lesson 2 55
Dangin saw the genitive in Sumerian as being formed in two different ways: either by
simple "juxtaposition" of two nouns (lyw-uru, "king-city" = "king of the city"), or by an
ending .a of a "general indirect case". He thought that the Ikl which appears when a vowel
follows the genitive marker was "inorganic"; it was a "hiatus-breaker" to avoid a sequence
of two vowels. Poebel effectively destroyed Thureau-Dangin's views, but traces of the
latter are still encountered in some works. Poebel' s work was further elaborated by
Jacobsen (1973).
- Typology
Scholars have pointed out previously that the genitive in Sumerian behaves differently
than the (other) cases. This is typical of agglutinative languages, where the genitive will be
cumul ative with respect to cases.
In most S-O- V languages, genitive constructions are expressed by the sequence
"possessor-possessed" (regardless of the exact morphological devices used). Sumerian,
then, would seem to be atypical, in that the sequence is "possessed-possessor". It will be
seen later that Sumerian also possesses a genitive construction of the type possessor
possessed, but this construction is not as common as the possessed-possessor one.
In Sumerian, modifiers of nouns typically follow their head noun. In this lesson an
adjective follows its head noun, and in Lesson 7 a relative clause follows its head noun. It
is a general characteristic of S-O- V languages for modifiers to follow their head nouns.
- Proper names
Many different etymologies of Ki-en-gi have been proposed, and just as many ex
planations for the derivation of Sumeru from Ki-en-gi. The sheer variety of such expla
nations shows how unsure such attempts are. Some of those proposed by more prominent
Sumerologists incl ude:
Anton Deimel: Ki-en-giCI) = ki.gir "land of the foot", i.e. "stopping place".
Edmond Sollberger: Sumeru is the Emesal form of Ki-en-gi(r), whatever the ety
mology of the latter might be.
Many other dubious etymologies have been proposed. They illustrate the fact that
there is really very little evidence to make a positive decision; the data can be made to fit
many different interpretations.
56 Lesson 2
-Titulature
nitab-kalag-g!! is a very old title, attested even with rulers preceding the Akkad dynas
ty. It was also used by Utu-hengal of Uruk. It is difficult to say exactly what an expres
sion like "strong man" or "mighty man" means; Hallo says: "'strong man' (that is, we
might almost say, independent ruler)" ( 1 966: 1 38).
Ur-Nammu was the first Mesopotamian ruler to use the title l!!gal Ki-en-gi Ki-uri. It
was used by his son Shulgi, but not (apparently) by the other rulers of the Ur III Dynasty.
It was used sporadically by later rulers (in both a Sumerian and an Akkadian form), right
down to the Persian period, especially by conquerors of Babylonia (such as Cyrus).
Ur-Nammu first assumed this title about the fourth year of his rule. In the early years
of his reign, the extent of his control was too limited, and his hold too weak, to permit use
of such a grandiose title.
Lesson 3
This brick is on display in the British Museum. The display stand was obviously
made many years ago. It refers to Ur-Nammu as "Ur-Gur", it refers to Nanna as "Nannar",
and it dates the brick to "about B.C. 2500" - about four centuries earlier than today's
chronology.
57
58 Lesson 3
Notes
bAd The PSD translates bad as: "wall", "fortification". Its normal Akkadian equivalent is
duru, glossed by the CAD as: " 1 . city wall, fortification wall, 2. inner city wall, 3.
fortress, 4. enclosure of a house".
Lesson 3 59
Text 3 a
* A) HI n o>
b>Jffl> If 7H
l'::f *
R=l 1=1
<V1:::1 Jfiil P>
60 Lesson 3
Text 3b
n p
" ii> t>
* lif
If
m
t>
Notes
Both inscriptions are inscribed on bricks, yet the signs in Text 3a are much more linear
than the signs in Text 3b. Because of the clearly wedge-shaped form of the signs, Text 3b
looks more like what we are accustomed to think of as cuneiform.
Differences between Text 3a and Text 3b also occur in lines 4 and 7. The various
signs which make up the geographical names are differently divided between the two lines
in each case.
Lesson 3 61
Discussion: structure
The basic structure of this text is:
- Sign fonnation
Some cuneifonn signs are, in origin, combinations of two different signs, one of
which is pictographic, and one of which is a phonetic indicator of some kind. For example,
the bad-sign is a picture of a city wall, with an inscribed bad-sign (); the function of the
inscribed bad-sig'n is to aid the reader in the correct pronunciation (another example occurs
in Lesson 15).
In Lesson 1, it was mentioned that the second element of the Nanna-sign (),which
62 Lesson 3
was interpreted by later Mesopotamian scribes as a ki-sign, may originally have been a na
sign, functioning as some kind of phonetic indicator.
- Co-ordination
- History
Most building inscriptions refer to only one undertaking (for example, the building of
a single temple), but it is not uncommon to find such inscriptions referring to two closely
related activities.
The wall referred to in this inscription was undoubtedly the wall which surrounded the
city of Ur. Woolley describes it as follows:
The walled city was in shape an irregular oval, measuring about 1 1 30 yards
in length by 750 yards in width, and was surrounded by a wall and rampart.
The rampart was of mud-brick with a steeply sloping outer face . ... Along the
top of this ran the wall proper, built of burnt bricks. ... Of Ur-Nammu's wall
not a trace remained ... just because the defences of Ur had been so strong
the victorious enemy [that is, the later Elamites who sacked the city] had
dismantled them systematically, leaving not one brick upon another
( 1 982: 1 37-8).
Although Ur-Nammu's wall may have originally enclosed the entire city of Ur, the city
rapidly expanded beyond these walls. The original walled city may have comprised no
more than one-fourth or one-fifth of the city in the Isin-Iarsa or Old Babylonian periods.
The situation must have been similar to that of many Middle Eastern cities today, with a
core consisting of an "Old City" (often referred to as a "madina" in Arabic or English), but
with much extension beyond it.
Lesson 3 63
Text 3 c
supplemen tary
The supplementary texts included here and in subsequent lessons are meant for prac
tice and review. They will nonnally contain no new vocabulary or grammar; any new
features will be explained.
This particular text is another brick.
rt& 'f
rf
Lesson 4
This inscription was engraved upon a stone bowl. No photograph is available.
Tf . . . J1 . ru
.. to dedicate a votive object
Notes
Nin-gal Ningal was the wife of Nanna, and the mother of Inanna. Being the wife of
Nanna, she was especially worshipped in Ur. Her name means "great lady".
65
66 Lesson 4
Text 4
* Er-
1T
P==r 8
:<l FrtlJ>
If ....
6 t
{J
ij la
f-11l1 $.
-j;
Tf )l
Lesson 4 67
5: .J!
I g- Ski-ma
I Unm lugal.Urim.a(k) the king of Ur,
7. Sumerian does not have many processes of word formation. However, it does have a
formative element nam. Prefixed to verbal and nominal roots, it produces what are called
"abstract nouns". For example, l!!gal is "king", nam-l!!gal is "kingship"; tU is "to live",
nam-tU is "life".
M is a case-marker not seen up till now, marking the "terminative" or "directive" case.
It does not have any one exact translation into English, although it generally indicates
"direction towards" or "action towards". Here, the meaning is something like "for the sake
of', or "on behalf of'. This phrase means something like "for his life", that is, "so that the
king will live a long time".
The exact phonemic value of this case-marker is not actually sure. Since the dative
case-marker Iral ) Irl after a vowel, one might expect Iel ) IU after a vowel. Such a
reduction does occasionally happen, although not apparently in this particular expression
(the sign rB does not appear to have any reading in liU). The original value of this
morpheme may have been, in fact, lee/. Conventional S umerological practice is to simply
transliterate it as e. It will be further discussed below.
The two signs transliterated here as: tU-la are also found transliterated as: !i-la. This
is the same problem seen earlier with kalag-g - kala-g, and UrimskCma - Uris ki-ma. In
this particular case, there is the added irritation that the sign l>-'(T is read !i (with no dia
critic) and til (with diacritic) .
8 . ...ffi, "to dedicate a votive offering". This i s a "compound verb", a type of word
formation very common in Sumerian. A compound verb is made up of two elements. The
first element of the compound verb immediately precedes a complete verbal form, including
its prefix chain. The second element functions as the regular Sumerian verbal root. In this
case, the first element is ; then comes a regular verbal form, consisting of a prefix chain
and the verbal root ru; ru is the second element of the compound verb.
The original function of the two elements of a compound verb is often not clear. In
many cases, the first element is (historically, at least) the patient of the verbal root which
forms the second element. For example, g!l... de means "to speak". de means "to pour out",
and g!l means "voice". Historically, then, this particular compound verb means "to pour
68 Lesson 4
out the voice"; that is, g!l is historically the patient of de. In such cases, the first or second
element may appear elsewhere as a free morpheme, functioning in other contexts like any
other noun or verb.
In other instances, the nominal element of the compound verb is in one of the adverbial
cases. For example, "to found" or "to establish" a temple or building of some kind is:
ki...gar. gm: means "to place"; ki means "ground". Here, ki is in the "locative" case (ki.a,
Lesson 6). The historical meaning was "to place (something) on the ground".
In some cases, the relationship between the nominal and the verbal components is not
sure. In this particular case, for example, ru probably means "to send" (judging from its
use in other contexts as a "simple", that is, non-compound verb), but it is not known what
the element means here.
Synchronically, it is not easy to define the term "compound verb". Certain verbs, for
instance, are almost always used with certain patients (cf. g!l... de above). Are these
compound verbs or not? It i difficult to say; it is not easy to produce a rigorous definition
of compound verbs. Compounds are basically identified on semantic criteria. If the
meaning of the compound is more than the sum of its parts - that is, if it is loosely an idiom
- then it is felt as a compound.
In the sign-lists and vocabularies in this book, compound verbs are indicated by the
use of three periods: ...ru.
In line 7, the terminative case-marker e occurs. According to what was said earlier
about the dimensional-prefixes, one might expect this e to be cross-referenced by a
dimensional-prefix. The dimensional-prefix which cross-references the terminative e is i;
it follows the datival dimensional-prefix in the prefix chain. Therefore, one might have ex
pected to find a form such as: -mu-na-i-(!l)-ru.
This verbal form illustrates the basic problem of the dimensional-prefixes. Although
theoretically there is a one-to-one correspondence between case relationships and
dimensional-prefixes, in practice it isn't so. Case relationships are found that are not appa
rently cross-referenced by a dimensional-prefix, and conversely dimensional-prefixes are
found even when no case relationship is apparently present.
This is probably more than just an orthographic problem. That is, it is not simply the
case that the dimensional-prefix is "there", but not written. In all the Ur III votive
inscriptions, the e in expressions of the type nam.til.anUe seems never to be resumed.
The presence or absence of dimensional-prefixes probably depends on semantic
factors at the level of the sentence, and at the level of the discourse, that is, beyond the level
of one single sentence. Certain nominal phrases are less closely bound to the sentence or to
the discourse than others. For example, nam.til.anUe is only loosely bound to the
sentence; it could be omitted, without any great loss of information. Such loosely-bound
phrases (almost idiomatic or formulaic in character) may perhaps not need to be resumed,
while such important constituents as the benefactive phrase would need to be resumed. Not
a great deal is known about the structure of Sumerian at this discourse level.
To sum up, although the problem is still open to discussion, it is more likely that it is a
semantic-syntactic problem, and not an orthographic problem. Therefore, no dimensional
prefix i is indicated in transcription.
Lesson 4 69
Discussion: structure
The structure of this text is:
nam is the regular element used to form abstract nouns in Sumerian. nig is regularly
used to form concrete nouns from verbal roots; this is discussed in Lesson 22.
- Phonology of -M
- Usage of -e
For anyone who has studied Akkadian, the Sumerian terminative in Iel immediately
calls to mind the Akkadian terminative in liU. To some degree, the two morphemes
overlap both in form and in function. At least as far back as 1 925, it was proposed that the
Akkadian terminative morpheme was borrowed from Sumerian (Albert S chott). However,
most scholars do not accept this view, for two reasons: First, the Akkadian terminative
appears to have cognates in other Semitic languages; it has recently turned up in Eblaite, in
certain limited contexts. Second, there do not appear to be any cases where Akkadian has
borrowed Sumerian grammatical morphemes. Similarly, it is difficult to believe that the
Sumerians borrowed the morpheme from the Akkadians. Although rare, it does show up
in very early Sumerian texts.
It is more likely that the two are independent developments. It is not impossible,
however, that the chance formal similarity between the two morphemes has caused the two
to influence each other in meaning, pulling them closer together in meaning than they may
have been at some earlier period. Given the fact that Sumerian and Akkadian were in close
contact for over a millennium, such reciprocal influence upon the grammar is not too
surprising.
- Compound verbs
Most compound verbs are of the type noun-verb, where the noun is (historically) the
patient of the verb. It is not sure whether these nouns are to be regarded as patients in
synchronic terms; this problem is discussed in Lesson 1 2. In some cases, the noun is in
one of the adverbial cases. However, more complicated compound verbs also occur, of
differing types: adjective-verb, noun-adjective-verb, and even noun-noun-verb. In the case
of the latter, one noun is (historically) the patient, and the other is (historically) in an
adverbial case. An example is "to pray", kiri3 .u ... W, literally, "to place (g@ the hand
(]y) on the nose (kiri3 )"; kiri3 is either in the locative case (kiri.a) or the locative-terminative
case (kiri.e).
Because there is no obvious formal way to define compound verbs, it is a legitimate
question to ask whether such a class of words actually exists. If more were known about
the etymology of each individual case, one might be less inclined to even posit the existence
of the class of compound verbs.
- Votive inscriptions
life of the king (the donor might or might not b e the king himself). The objects were not
"functional", as we would understand the term; that is, this bowl was not actually used as a
daily eating utensil. Similarly, the votive cylinder-seal in Text 22 was probably not used as
a daily, routine, cylinder-seal.
Except for the component of the compound verb, there is no direct object (patient) in
the text. This is because the votive object itself can be thought of as constituting the direct
object (patient).
The use of the term "votive" to describe such inscriptions has been criticized by A.
Grayson:
The etymology of the word "votive" implies a vow and, since no vow is
involved in the ancient Mesopotamian texts under discussion, the term is
incorrect. They are certainly not "votive" or "ex-voto" inscriptions in the
ancient Roman sense where a vow preceded the dedication ( 1980: 1 57).
For this reason, Grayson prefers to refer to such texts as "dedicatory inscriptions".
72 Lesson 4
Text 4a
supplemen tary
Lesson 5
t::f ka
Notes
En-IU Father of Nanna; the most important god of the Sumerian pantheon. His name
means "Lord Air" or "Lord Wind", but he was in general responsible for the orderly
running of the universe (although he had a destructive side as well).
Enlil functions as the active leader of the Sumerian gods; he has apparently displaced
the sky-god An from this role. He was worshipped at many places, but his special
sanctury was the E.-kur in Nippur.
The lil-sign is the same ke4 -sign seen previously.
(The term en, "lord", is discussed in Lesson 9.)
k:ur The original meaning of this word was probably "mountain"; the kur-sign, in fact, is
thought to be the picture of three mountain tops. The word then comes to mean "foreign
land".
something like "water coming from the sweet-waters", as opposed to "water coming from
the salt-waters", and as opposed to "rain" (Mg, which is composed of the -sign followed
by the an-sign: TT ).
nidba The reading is uncertain; it is also transliterated as nindaba. When scholars are
unsure of the reading of a "compound logogram" (a single "word" graphically composed of
several individual logograms), they occasionally add in parentheses the reading of the
component parts. Thus, this sign is also transliterated as: nidba (PAD-dINANNA), or any
of several variants, such as: nidba (SUKUR2 _ dINANNA), since it is in fact not clear
exactly what all the components of this particular logogram are!
This word was borrowed into Akkadian as nindabu, although the forms nindabbu,
nigdabbu. nidabu. nidbu and nidpu also occur. CAD translates the Akkadian word as:
"cereal offering, food offering, provisions".
ba-al This is the normal verb used to describe the restoration of a canal, that is, clearing it
by dredging it of accumulated silt and trash.
The verb is almost always written in this way, with two signs. There are a few
instances where it is spelled bal or ba-la. It is not sure what the writing ba-al implies about
Sumerian phonetics. The vowel may have been long, or there may have been a glottal stop
or a glide between the two / a /-quality vowels. Because of this unusual writing, it has even
been speculated that the word is a borrowing from an as-yet unidentified language. For
convenience sake, it will be transcribed here as: ba-al.
Lesson 5 75
Text 5
Jb4
p ,
f-J( !l
$.t
\1ijg W
1 ii
yY Jf \ p.- -r
Yf ttr ,O
rw
76 Lesson 5
Discussion: structure
The structure of this text is:
- Writing system
The word for "food offering" illustrates a problem not yet seen in the script. Its
pronunciation as Inidbal is given by various lexical lists, where it is spelled out syllabically
as: ni-id-ba. It is thought to derive from ninda "bread, food" and ba "to divide, to ap
portion"; ninda would presumably be an incorporated direct object (patient) of ba. (It is
occasionally spelled ninda-ba.) The pronunciation as Inidba/ shows a phonetic reduction
of */nindabal > Inidba/. The various spellings in Akkadian, mentioned above, reflect both
the older and later S umerian pronunciations.
It is difficult, however, to say how this word came to be written as "PAD
dINANNA". That is, there does not appear to be any way to phonetically relate the word
Inindabal or Inidbal to the individual signs forming this logogram. Rather, the reading
Inidbal refers to the "sum" of the three signs. Without the evidence of lexical lists, in fact,
there would probably be no way to figure out that this group of three signs was to be read
as Inindabal or Inidba/. The pad-sign has several different readings, and in Akkadian
stands for several different words: kusapu, "a kind of bread" (probably of Semitic
etymology); kurummatu, "food portion" (a Sumerian word), etc. The significance of the
Inanna-sign (if that is how it is to be understood here) is uncertain.
- Proper names
In the bilingual texts from Ebla, the equivalent of En-lil is given as: I-li-Iu. This
seems to agree with later Akkadian pronunciations of the name, which also show an
assimilation of IEnlill > IIllil/; some Sumerologists, in fact, transliterate the two signs En
lil together as: Elli!. (It has also been speculated that the interpretation of the name as
"Lord Wind" is a Sumerian folk-etymology, and that the word is of pre-Sumerian
etymology.)
- Cones
The building inscriptions seen up to this point have all been inscribed on bricks. Text
5, on the other hand, was inscribed on what is commonly known as a "clay cone" or "clay
nail". Clay cones were used throughout Mesopotamian history; their form and function
varied to some degree from period to period. A detailed description is that of Grayson,
describing the clay cones of the Neo-Assyrian period. The clay cone
is an oblong conical object of clay. It is tapered almost to a point at one end
and at the other there is a large semi-spherical head. The same inscription
usually appears on both the shaft and head. The shaft was commonly
inserted in the upper portions of walls with the head, which was painted a
bright colour, protruding ( 1 980: 145).
In this view, the clay cone would have been at least partially visible to on-Iookers.
Other scholars belive, however, that the protruding end would have been plastered over,
covering up the inscription (at least, in the Ur III period). In fact, the exact purpose of these
cones is still something of a mystery to us.
Woolley found such cones in situ, forming part of the terrace of the ziggurat of Ur-
78 Lesson 5
Nammu:
Such cones were familiar enough as objects on museum shelves, but now for
the first time we saw them in position just as the builders had set them four
thousand years before. ... One felt a quile unscientific thrill at seeing those
ordered rows of cream-coloured knobs which even the people of Ur had not
seen when once the tenaced wall was finished and plastered (1982: 140).
The latest such cones found in Ur date from the Neo-Assyrian period. Curiously, they
were not in the wal1, but were buried below the floor.
Clay cones were usually inscribed in several duplicates; at least eight copics ofTcxt 5
are known. The point of such cones, again quoting Woolley, was not to "parade [the
ruler's] achievements before his fellow-men, but to keep the record of his piety fresh in the
mind of the god, who presumably can see through a brick wall" (1982:228).
Some scholars differentiate between clay "nails" and clay "cones". Gelb says that clay
nails "are easily recognized by their mushroom shape. with broad, thick heads and short
shafts. ... Clay cones are characterized by a total or almost total lack of the head"
(1948:267). Different places and periods seem to prefer one or the mher; it is also possible
they had different functions. Gelb says that "the function of such nails and cones is much
the same as that of tablets commemorating the erection of public SlI"Uctures in modem
times" (1948:268).
No photograph of Text 5 is available. The fOllowing is a photograph of a dedicatory
cone from the rime of Gudea's father:
It was mentioned in Lesson 3 that it is not too uncommon for building inscriptions to
treat more than one activity. In the case of building inscriptions inscribed on clay cones, it
is less common for more than one acriyity to be mentioned. In this particular text, the
Enerinnun canal may have brought the waters which irrigated the fields of the temple being
rebuilt.
All of the Ur HI kings were involved with repairing the canals and drainage systems
of Mesopotamia, by dredging and rebuilding. The names of many canals in use during the
Ur ID period are known; most of the names are Sumerian. Urnammu is known to have
built and to have repaired a number of canals. It is usually assumed that much damage had
been done to these systems by the Guti, who are thought to be responsible for bringing
about the fall of the Dynasty of Akkad. Revisionist thinking, however, believes that the
Guti did not do as much damage as is commonly thought, nor did they hold that much
control over Mesopotamia. Most scholars seem to feel that inscriptions such as Text 5
refer more to routine maintenance and expansion of the canals; various kinds of
administrative texts refer to such activity.
Lesson 6
This is another brick of Ur-Nammu.
An An (DN, masc)
%- digir god
1- barag (bar) dais
ki place, earth
Notes
gi In addition to its use as a determinative, gi means "tree; wood; object made of wood".
The Akkadian equivalent, i1!, is translated by the CAD as: " 1 . tree, 2. timber, lumber,
wood, wooden implements, aromatic wood, firewood, 3. wooded area". It is sometimes
transliterated as ge.
An The god of the sky. At one time he may have been the active leader of the S umerian
gods, but at some point prior to our written records he was displaced in this role by Enlil.
Scholars sometimes refer to him as "shadowy", or as a kind of deus otiosus.
His name is almost always written without the divine determinative.
digir It is usually assumed nowadays that this word was pronounced with a velar nasal.
Some think it may have been pronounced I dingir/, and in fact it is most commonly
transliterated as dingir.
79
80 Lesson 6
gBkiri6 This is also transliterated with an initial / g/, especially in older transliterations:
giri 1 1 and giri 1 2 . Some Sumerologists do not think that the gi-component here is a
determinative, and so it is also transliterated: gi-kiri6 . (This is also the view found in
some lexical texts.) However, other Sumerologists believe that the kiri 6 -sign includes what
here is called a determinative; that is, the one sign kiri6 is composed of two separate ele
ments: Cl' and.mr:J , and so there is no determinative. The most current practice is to
read the two signs as gikiri 6 .
barag The PSD translates this simply as "dais"; many Sumerologists translate it as
"throne-dais". It was borrowed into Akkadian as parakku , which is translated by AHw as:
"Kultsockel, Heiligtum".
Text 6 was inscribed on a brick which apparently was part of the barag of the temple.
mal} This is not common in finite verbal forms. The most common Akkadian equivalent
of its use as an adjective is Trl1, translated by the CAD as: "first-rank (in importance,
quality), outstanding (in size), august, excellent (used only as a poetic term)".
sikil This is also not common in finite verbal forms, but its adjective is frequent. The
usual Akkadian equivalent of the adjective is ellu, glossed by the CAD as: " 1 . clean, pure,
2. holy, sacred, 3. free, noble".
Lesson 6 81
Text 6
82 Lesson 6
inscriptions, the locative .a is usually not cross-referenced. This is probably more than just
an orthographic problem. In Lesson 4, an instance occurred where the dimensional-prefix
for the terminative was not expressed. There, it was said that it may have been because of
the idiomatic nature of the phrase. However, such an explanation hardly fits the numerous
cases where the locative case is not cross-referenced. Perhaps locative phrases in general
were felt as less closely bound to the verbal phrase than were the other adverbial cases.
The logical antecedent of the datival dimensional-prefix .na is the nominal phrase in
lines 1 -2 of this inscription, which is part of a different, independent sentence. The same
situation occured in Text 3 , where a datival dimensional-prefix referred logically to a
nominal phrase in a preceding sentence.
Discussion: structure
The structure of this text is:
-Amissability
It may be useful here to summarize the ways the amissable consonants are reflected in
the writing system. In the case of grammatical morphemes (which are normally written
syllabically), such as the genitive .ak, the amissable /k/ does not show up in word-final
l
position: "King of Ur" is written: lugal-Urim t ma. When not in word-final position, it
does show up: The same expression, with an ergative case-marker, is written: l!Jgal-
Unm ki-ma-cl4.k
S
In the case of lexical morphemes (which are normally written logographically), it
cannot be determined from the writing system whether the Auslaut was pronounced or not.
That is, just by looking at the sign, there is no way to tell whether the word for "dais" was
read /barag/ or /bara/.
When the amissable consonant of a lexical morpheme is not word-final, there is less of
a problem in understanding the phonology, but still a problem in understanding the nature
of the orthography. As discussed in Lessons 2 and 4, should a form such as "for the sake
of his life" be understood as nam-ti I-la-ni-e or as nam-!l.-la-ni-se? Or should the adjective
for "mighty" be understood as kalag-g.a or as kala-gf!? Phonetically, probably all Sume
rologists would understand these forms to be /namtilange/ and /kalaga/. The question,
rather, is how does Sumerian orthography represent these pronunciations.
- Loan words
In a number of early loan words from Sumerian into Akkadian, final voiced stops
appear as unvoiced, and usually as geminated: isib ("kind of priest", Lesson 19) ) mppu;
Mrig ("an official") ) abarakku. Word-initial voiced stops usually appear as unvoiced in
84 Lesson 6
-Locative
Locative phrases such as "in the earth", ki.a, are almost always written ki-. That is,
the lal of the locative case-marker almost always appears in the writing. It has been
speculated that the lal of the locative case-marker actually assimilates into or contracts into
a preceding vowel, but the script is morphographemic, and writes the lal anyway.
The locative case-marker .a does not usually appear when the head noun is the first
element of a compound verb (kiri 3 .u ... gal was mentioned in Lesson 4). It is possible that
this is more than a case of assimilation or contraction. As discussed in Lesson 4, it is not
sure how "present" case-markers were in the case of compound verbs, in a synchronic
sense; they may have been deleted or somehow reinterpreted.
Although in the Ur III royal inscriptions the locative case is usually not cross
referenced by a dimensional-prefix, in the contemporaneous Ur III administrative docu
ments it quite frequently is cross-referenced. As was mentioned in Lesson 4, this may
mean that factors such as genre and style must be taken into account, to describe and
explain the different distributions.
-Adjectives
In Text 6 the adjectives man and sikil occur. In Text 2 the adjective kalag-g occurred.
As stated above, some adjectives end in .0, others in .a. Some adjectives are occasionally
found sometimes in .0, other times in .a. This situation is not well understood. The most
recent discussion is by Joachim Krecher ( 1 978). He believes that, at least in certain cases,
the forms in .a mark a nominal phrase as "definite" or "determined" in some way; those in
.0 are the unmarked forms.
The difficulty in investigating this problem (and many other problems in Sumerian
grammar) is that it is not easy to find sentences which are very close in structure, but differ
only in the presence or absence of .a on an adjective. There are usually too many variables
involved, to be able to sort them all out.
Lesson 7
Two copies of this inscription are reproduced. The variation in the shape of the
cuneifonn signs is fairly minimal. Text 7a is a stone foundation tablet. Text 7b is a brick.
The inscription appearing on them is a standard inscription, a further sub-class of royal
inscription as distinguished by Hallo.
1t1 man
Notes
Id According to Gelb, "The S umerian word 1t1 is a noun meaning 'person, man' (in the
sense of homo, Mensch, not vir, Mann), and may be used for both males and females"
( 1 979b:5 1). Jacobsen says that 1t1 "denotes a man (Akkadian awIlum) or woman (Ak
kadian aWIltum) who heads a household, firm, or city" ( 1 987: 1 30 n. 17).
85
86 Less<!n 7
Text 7a
Lesson 7 87
Text 7b
88 Lesson 7
consonant, this conjugation-prefix is nonnally written by the i-sign (that is, the sign which
occurred earlier with the value ni).
Second, here the personal-affix 1nl cross-referencing the agent is expressed in the
script. The in-sign expresses the combination of the .i conjugation-prefix with the .n of the
person al-affix. This particular use of the in-sign is quite regular.
It is difficult to say why the finite verb fonns seen previously use the conjugation
prefix !lli!, but the nominalized fonn uses the conjugation-prefix i. As discussed in Lesson
1, the essential difference between mu and i is elusive. Jacobsen says that the conjugation
prefix i "presents the occurrence denoted by the verb as touching on the subject without
inwardly conditioning him in any lasting manner" ( 1 965:25 1). mu is the
mark of location of the occurrence denoted by the verb on the inside border
(.u) of the area of the speech situation (m.) . ... It adds to this implications of
emotional involvement of the speaker, of his being personally engaged
( 1965:254).
However, not all scholars are as certain as Jacobsen in their conviction; Poebel, for
instance, seemed to have the opposite view of the relationship between mu and i. J.N.
Postgate has expressed perhaps the most negative view:
For many years a vexed question in Sumerian has been the distinction
between the prefixes mu- and i-, and our failure to define the difference in a
satisfactory way has epitomized our helplessness before Sumerian grammar
as a whole ( 1974:24).
And, as discussed in Lesson 1, in actual practice the conjugation-prefixes are basically
ignored in translation.
It is likewise difficult to say why the personal-affix 1nl is written here, whereas it did
not appear in any of the previous finite verbal forms. In the previous texts, the finite verb
fonn always had an expressed subject (agent). In this particular text, there is no expressed
agent, since the verb is inside a relative clause; the h.i of line 4 is the relative marker, not an
agent of the sentence as a whole. One might hypothesize a rule such that "expressed agents
do not use the personal-affix 1nl, but verbal forms without expressed agents do", but from
other texts it is known that the situation is not nearly as simple as this.
To sum up, III of line 4 is the relative marker of the relative clause. The relative clause
is fonned by nominalizing a finite sentence, by use of the nominalizer .a. The nominalized
clause stands in apposition to the relative marker.
Discussion: orthography
In both copies of this text, the relative clause is split into two cases. The first case
contains the more nominal component, and the second contains the more verbal component.
It is not uncommon for long relative clauses to be split into two or even more cases.
The S umerian word for "king", illgal, is a noun-adjective compound from Ill, "man",
and gal, "great, big". In older forms of the illgal-sign, the gal-component was written a
little above and to the right of the Ill-component. At times the two signs were totally
separated, and can even be written on two different lines of one case. As the cuneifonn
signs gradually became more linear, the gal-sign shifted position, and so in "standard"
90 Lesson 7
Sumerian the illgal-sign is all one sign, with the gal-component in front of the hl
component.
As will be discussed in Lesson 1 3, the order of cuneiform signs within a case in the
earliest texts was to some degree free, with the order-as-written not necessarily reflecting
the order-as-read.
- Conjugation-prefixes
As hinted at above, the fact that the conjugation-prefix 1. appears in the relative clause
instead of mu raises several obvious questions: Does 1. appear outside of relative clauses?
Does mu appear inside of relative clauses? What about the distribution of 1. and mu in
general? What about different types of relative clauses: clauses where the relative clause
modifies an agent ("the man who built"), clauses where the relative clause modifies a
patient ("the house which the man built"), etc. What about the distribution of the
conj ugation-prefixes in topic ali zed and emphatic sen tences?
Unfortunately, not all of these questions can be answered. The data are both
ambiguous and limited; the number of attested relative clauses is not that large. Without
access to native speakers, such questions cannot be tested.
In the Ur III texts, the verb form in the relative clause in this particular expression
seems always to appear as: in-du-. In some bricks of Gudea, the form mu-na-du is used
in a sentence as a main verb, while the verb of an embedded relative clause uses the form
in-du-:
dNin-gi-zid-da digiI-ra-ni
Gu-de- ensi2 Laki III E-ninnu dNin-gir-su-ka in-du-
Gir-suki-ka-ni
mu-na-du.
- Standard inscriptions
It will be noticed that there is no finite verbal fonn (or any other predicate) in this text!
Instead, there is only a personal name, followed by a series of titles or epithets. Jacobsen
has referred to such sentences as "label sentences". Hallo, in his discussion of the different
categories of royal inscriptions, refers to such inscriptions as standard inscriptions or
property inscriptions. By "standard", Hallo means that the text is something like a flag or
other identifying device: It identifies the building as being the property ofUr-Nammu.
Such inscriptions, which serve to indicate ownership of a building, are often regarded
as the "simplest" fonn of royal inscriptions. They usually consist of a royal name,
followed by a limited number of epithets, one of which may be a relative clause, as in Text
7a/b. There is no verbal predicate, These texts can be very short; Text 1 3c is a standard
inscription consisting of only two lines.
- Foundation deposits
The building inscriptions presented here up to now have been inscribed on either
bricks or clay cones. Text 7b is such a brick. A third category of building inscriptions is
referred to as "foundation deposits". These were actually buried under the foundations of
walls in a building, in a small pit. Text 7a is one sub-type of such foundation deposit, a
stone foundation tablet. In Woolley's words:
Foundation-deposits are found in the corners of buildings. Built into the
wall-foundations there is a small box of burnt bricks, lined with matting and
waterproofed with bitumen; in it is set a copper figure of the king modestly
represented as a labourer carrying on his head a basket of mortar; at his feet
is a stone tablet in the fann of a pIano-convex brick; on the brick and on the
king' s skirt is an inscription recording his name and that of the temple
( 1982: 1 61).
Occasionally, the figurine or the tablet is uninscribed. It is thought that these
foundation tablets were intended to represent "model bricks".
The following photograph is of a figurine which was found in Nippur along with Text
7a, and so bearing the same inscription. (At least two of these bronze canephore figurines
are known.) The figurine is thought to represent Ur-Nammu himself, carrying a basket on
his head with the building materials used to make the "first brick" of a building.
r.esson 7
92
Lesson 7 93
Text 7c
supplementary
Another brick.
Lesson 8
This text was inscribed on what is commonly called either a door socket or pivot
stone.
ki...i!gi!2 to love
Notes
E -kur This was the main temple of Enlil in Nippur, by far his most important sanctuary.
It was in the Ekur that the assembly of the high gods would meet, as occasion demanded.
The god An presided over these meetings, but it was the responsibility of Enlil to carry out
the decisions. One of the reasons meetings were held was to select the rulers of Meso
potamia.
95
96 Lesson S
Text 8
Ef;>- . FlI
l fJ =-
--... * rEI
5 4f
I!ffir=> ,.-JJ
" ID) ;lmEJ
==1
d
JO H >-
Lesson 8 97
Discussion: structure
The structure of this text is:
- Door sockets
Many texts inscribed on door sockets have been preserved; for a photograph of such a
stone, see Text 17a below. Such stones were partially underground, and were used to hold
a door. In Woolley's words,
The Sumerian door consisted of a wooden leaf fixed to a pole rather higher
than itself; the projecting top end was held by and revolved in a metal ring
attached to the lintel, the lower end was shod with metal and went down
through a hole in the pavement to rest and turn on the hinge-stone. This was
98 Lesson S
Text 8 a
supplementary
Another brick.
Lesson 9
OO <Hf
E-temen-ni -guru3 Etemenniguru (TN)
en lord
g4 to return, to restore
bi, be
Notes
Unug One of the most important cities in southern Mesopotamia; it often played a role in
political history. Before becoming king, Ur-Nammu had been military governor of Ur
(akkana-Urim5kCma) under the control of Utu-Hengal in Uruk.
The etymology of the name is unknown; this is discussed below. The pronunciation
of the name as IUnugl is known from syllabic writings. However, Semitic spellings show
Ir/: Biblical Erech, modern-day Arabic Warka, etc.; the lexical lists also give the Akkadian
equivalent with Ir/ : U-ru-uk. It is not known why the Sumerian form shows 1nl while
the Semitic forms show Ir/.
sag This is literally "head". However, it not infrequently forms the second element of
noun-noun compounds, where, as in many languages, it can take on metaphorical uses.
For example, ka-, "beer-head", is "top quality beer"; dumu-, "son-head", means
"eldest son".
has several Akkadian equivalents: reu, "head" and several derived meanings;
q.!!qqadu, also "head" and several derived meanings; amelu, "man", lu, "young man", etc.
en This is normally translated "lord", a purely conventional translation. The ruler of Uruk
is always called an en. Its meaning is further discussed below.
101
102 Lesson 9
Text 9
Lesson 9 103
characteristics of the terminative case. In this particular fixed expression, the force
conveyed by the locative-terrninative is: "He returned the temple to its original place". That
is, its meaning here is close to that of the terminative marked by -e.
The form of the locative-terminative dimensional-prefix is much open to dispute. It is
often not cross-referenced at all in the verbal prefix chain. Other times, it seems to use the
dimensional-prefix corresponding to that of the locative case, that is, .ni. (According to
some Sumerologists, .ni in such contexts is for *n-e; the .e is the dimensional-prefix of the
locative-terminative, and the .n is an optional pronominal-prefix, discussed in Lesson 1 5.)
In this particular case, it is not cross-referenced at all in the verbal chain.
The .0 cross-references the patient. There is no expressed patient in this sentence; the
logical patient is the nominal phrase in line 1 1 . Similarly, the logical antecedent of the
datival dimensional-prefix .na is the datival nominal phrase in lines 1 -4.
Discussion: structure
The structure of this text is:
- Sign formation
Over time, the Sumerian writing system produced a number of new cuneiform signs;
the history of these developments is rarely visible to us. One such process was the addition
of short strokes to an already existing cuneiform sign, thereby modifying its meaning. For
example, the word sag, "head", is represented by a sign which was originally the picture of
the head and upper torso of a man: x:..CY::-f ' The word ka, "mouth", is represented by the
same sign, but with the addition of short strokes over the region of the mouth: lf...:r .
Akkadian scribes referred to these extra strokes as gunO. This is an adaptation of a
Sumerian word gunu, apparently meaning "colored". Modern scholars will sometimes
refer to, e.g., the ka-sign as "SAG+gunG". This convention is sometimes necessary, when
the value of a "gunOfied" sign is not known.
- Functions of .e
The locative-terrninative case is marked in .e, and the agentive-ergative case is marked
in .e. An obvious question which springs to mind is, are the two related? It is probably no
accident that the two case-endings share the same phonological shape, / e/. The parallel has
been made to the English preposition "by". This can express a locative ("by the river"), an
instrumental ("by the hammer"), or an agent ("by the man"). The agentive marker .e may
have developed out of this locative-terminative .e. This .e may have started to lose some of
its functions, which began to be taken over by the locative case in .a and by the terminative
Lesson 9 105
case in .e. Haayer has discussed this point, in terms of universal tendencies in language:
One of the most characteristic features found in case marking in ergative
languages is that the ergative case is often identical with another case, most
often the genitive or instrumental, sometimes the locative or dative. In
Sumerian, for instance, the ergative case is marked by the postposition -e,
which is identical to the locative-terminative -e, and is in origin a deictic
pronoun ( 1986:80).
Although we understand and translate simple Sumerian sentences such as lugal.e e.0
mu.n.du.0 as "the king built the house", it is usually assumed that at some "Proto
Sumerian" stage, the meaning may have been something like "a house got built, connected
with the king", or "there was a building of a house by the king". That is, to some degree,
pre-historic Sumerian (and historic Sumerian?) should be understood as basically "passive"
in nature. It has been said that in ergative languages, the patient of the sentence is the
"topic" of the sentence (while the agent is the "comment"), but in accusative languages, it is
the subject which is the topic (while the patient is the comment).
- Noun compounds
- History
The different functions of the en and l!!g al have been much discussed; they varied to
some degree from place to place and from period to period. In Jacobsen's seminal article on
"Early Political Development in Mesopotamia", he stated that in the earlier periods the en
(Akkadian blu) was more of an "administrator" while the l!!g al (Akkadian arru) was a
"warleader":
In the case of the en the political side of the office is clearly secondary to the
cult function. The en's basic responsibility is toward fertility and abundance .
... The "king", l!!g al, in contrast to the en was from the beginning a purely
secular political figure, a "warleader" ( 1 957 :375 n.32).
As for the more "original" meanings of the telms,
The Sumerian term en which is generally translated "lord" denotes basically a
productive manager, someone with magic gifts to make things thrive
( 1 987:20 n.2; cf. p. 277) . ... Under the early political forms ... the king
(l!!gil)l was usually a young man whose task it was to lead the army in war
( 1 987:236 n.4).
106 Lesson 9
- Substrate
The etymology of the name Uruk is unknown. Many of the oldest cities in Sumer
have names which are not apparently Sumerian. Such names go back to the language(s)
spoken by the people(s) living in southern Mesopotamia before the Sumerians arrived; the
name of the city of Ur may be one of these names. Gelb says:
Almost all the Mesopotamian geographical names found in the earliest
Sumerian sources are non-Sumerian and non-Akkadian and must be
assigned to the proto-population of Mesopotamia. This conclusion is true of
the names of rivers and mountains, as well as of cities and countries. Only in
the Pre-Sargonic period do we find the first attestation of Sumerian
geographical names ( 1 962:49).
(A number of these place-names are discussed in Limet 1 97 5b.)
These substrate peoples, about whom virtually nothing is known, were referred to by
Benno Landsberger as "Proto-Euphrateans"; they are also sometimes referred to as "Proto
Tigridians". The words for certain material objects and certain professions in Sumerian go
back to this language, for example, nagar, "carpenter", which has no obvious Sumerian
etymology. Some of these substrate words then passed on to Akkadian, and eventually on
to Aramaic, Hebrew, and Arabic.
Scholars disagree in their views as to how much of Sumerian vocabulary is of
substrate origin. Some are inclined to see a large number of substrate words in Sumerian,
including many place-names and divine-names; other scholars are less convinced.
Jacobsen, for instance, has Sumerian etymologies for several city-names, which other
scholars regard as substrate names. There is rarely enough evidence to decide any
particular case.
- History
- History
It i s difficult to say how literally the expression "he restored it to its place" should be
understood. The Sumerian phrase (and the corresponding Akkadian phrase) is somewhat
ambiguous; it can mean either "to restore to a former spot" or "to restore to a former state".
Kings of the Neo-Babylonian period specifically claimed that they took care to rebuild
temples exactly on old foundations. Woolley says that
It was customary in Mesopotamia, when rebuilding a temple, to incorporate
the earlier one within the core of the platform upon which its successor was
to be set. This often meant largely dismantling it ( 1 982: 1 09).
Lessoll 9 107
Nabonidus. for example. has left several inscriplions in Ur, in which he claims to have
restored the ziggurat of Ur-Nammu (the Eternenniguru). He states. in fact. that Ur-Nammu
started the work on the ziggural, but did not finish it; Ur-Nammu's son and suc",essor
Shulgi also worked on the complex, but did not finish it; only he. Nabonidus himself,
completely finished and restored it. The following drawing is WooJley's reconstruction of
Nabonidus' ziggurat; it is instructive to compare i t with Woolley's reconstruction of Ur
Nammu's ziggurat, given in Lesson 1 . Ur-Nammu's ziggurat itself was built over an earlier
temple, which itself was built over an even earlier temple.
....,. PO 8.C
_ . - .. ...
108 Lesson 9
Notes
S ul-gi In older transliterations the name was read as Dun-gi. It is almost always read Sul
gi nowadays, although there is really very little evidence to permit a decision one way or the
other. It is usually interpreted as "noble (gir) young man (ul)".
The gi-element had an Ir/-Auslaut, although the standard sign-lists do not record any
value in Igir/. Because of this Ir/-Auslaut, he is occasionally referred to as "Shulgir".
This gi-element may be the same gi-element seen in the GN Ki-en-gi.
an-ub The analysis is not clear. Because this word is occasionally written without the an
sign, some think that the an-sign is the divine determinative, and so it is sometimes
transliterated as: dub. Sollberger, for example, explains the word as: "part of the world
(as an emanation of the divine, hence the (divine) classifier)". However, these omissions of
the an-sign occur only in relatively late texts. For simplicity sake, it is transliterated here as:
an-ub, and transcribed as: anub.
1 09
1 10 Lesson 10
Some read the sign as limu2 instead of limmu2 . In older transliterations, it is fre
quently transliterated as tabtab. This is still preferred by some modern-day Sumerologists.
Text 1 0
t7
m II
* fHlli
tf f fIrt>
t=C] 5
< 8
*-
rt:r
Commentary
1 -2. The text begins with a benefactive to a goddess, even though there is no votive object.
The use of the benefactive is due to the formulaic nature of these inscriptions, coupled with
the idea that all official activity, including the regulation of weights and measures, was
conducted ad maiorem deorum gloriam.
3. At some point in the middle of his reign, Shulgi's name begins to appear with the
determinative normally reserved for divine names, *" . This (and other) evidence indicates
that Shulgi was deified, both during and after his life-time; this is discussed further below.
6. There are two genitive constructions in this line, one of which has not been seen
previously: the "anticipatory genitive". Up to this point, to express "the temple of the
king" in Sumerian, one would expect e.lugal.a(k), written -h!gal-la. However, Sumerian
has an alternative genitive construction, which puts the noun with the genitive marker first,
followed by the second noun with a possessive-suffix. A literal translation of this
construction would be: "of the king, his temple". This would be: lugal.a(k) e.ani, written
either lugal-la -i!-ni or lugill-la -ni.
This particular expression loosely translates as "the four quarters". A literal translation
would be: "of the corner-and-side, its four": anub.da.(k) limmu.bi. As in the expression
"king of Sumer and Akkad", there is no conjunction between "corner" and "side". Since
these terms are inanimate, the form .bi is used, not .ani.
However, this entire expression is itself the second element of a regular genitive
construction: "king (of the four quarters)". The first element in this genitive phrase is
h!gal. The second element of this genitive phrase is the entire phrase: anub.da.(k)
limmu.bi. The second element is then followed by the genitive marker .ak. The lil of the
possessive-suffix .bi contracts into the tal of .ak, producing Ibak/, and the Ik/, as is the
normal practice, is not written. Thus, a literal translation of this entire expression would be:
"king of [of the corner-and-side, its four] ": [lugal]. [anub.da.(k) limmu.bi] .a(k), producing
"king of the four quarters".
The anticipatory genitive tends to occur in certain fixed expressions (such as in line 6).
In theory, it can be used anywhere a regular genitive could be used, but in practice it is less
common. Since the expression "king of the four quarters" is quite frequent, it is not a
problem to recognize it in context. However, non-idiomatic uses of the anticipatory
genitive can be quite difficult to recognize. The two clues for its presence are: an otherwise
unexplained la/-vowel, followed a little later by an otherwise unexpected possessive
suffix. Several instances of the anticipatory genitive occur in the following texts.
Discussion: numbers
One of the lexical texts found at Ebla is a small tablet giving the names of the Sumerian
numerals from one to ten, spelled more-or-less syllabically. This tablet (TM.75.G.2 198)
,
was apparently some kind of school or practice text. For "four", the tablet says: li-mu,
presumably for Ilimmu/.
- Typology
It is more common for S-O- V languages to have a genitive construction of the type
112 Lcssoll lO
-Weights
Although stone was the usual material used for making weights, in Assyria bronze
weights were sometimes used instead.
Lesson 10 1 13
- History
The first Mesopotamian ruler to use the divine determinative before his name was
Naram-Sin, the fourth ruler of the Dynasty of Akkad, who ruled approximately 2254-22 1 8
BC. Gadd says that "no doubt the vast accession of power and width of sway won by
such a mighty figure as Naram-Sin helped to make him appear superhuman" ( 1 97 1 :619).
The determinative was used by all the Ur III rulers except Ur-Nammu. It was used only
sporadically by following rulers.
Occasionally, epithets in the royal inscriptions use the word "god". In Text 1 7a, for
example, Amar-Sin refers to himself as digIT-zid dUtu-kalam-ma-na, "the effective god, the
sun-god of his land".
There is also a certain amount of literary material which indicates that the Ur III kings
were considered, or considered themselves, "deified". However, it is not really known
what this means. The use of the English word "deified" is rather facile; it is very difficult to
say what this meant to the Ur III rulers or to their subjects. (Gadd says that "vainglory and
popular superstition supported [this policy]" [197 1 :6 19].) However, there is a certain
amount of evidence from royal tombs of the Ur III period to indicate that offerings were
made to the dead Ur III kings, implying that they were worshipped as gods after their
death.
Moorey says that
The most common evidence for the worship of the deceased kings of the Ur
III Dynasty is provided by economic documents describing deliveries to a
place called ki-fl-Mg, where liquid offerings to the dead were libated. No
thing specific is known of these mortuary shrines (1984: 17).
This topic of the deified king has been discussed by Jacobsen:
The deified king is not a "god" generally; he has the specific relation to the
country that a personal god has to his ward . ... The king, as leader of the
country and originator of policy, is the "personal god" of his realm. The
deification of rulers in Mesopotamia is accordingly to be understood not in
terms of the qualitative contrast human:divine, mortal:immortal, etc. , but in
terms of function of the king, he is the "genius" of the country ( 1 957: 395
n . 1 08).
Although the divine determinative is never used before Ur-Nammu's name in any of
his royal inscriptions, it is so used in the Prologue to the Law Code usually ascribed to him.
But this is probably the work of Shulgi, and in any case the Prologue is a rather late copy,
dating from the Old Babylonian period. Also in the Prologue, he is referred to as the son of
the goddess Ninsun; he is elsewhere referred to in the same way.
-Titulature
It was mentioned above that Naram-Sin of Akkad was the first Mesopotamian ruler to
use the divine determinative. He was also the first to use the title "king of the four
quarters", in both an Akkadian form and a Sumerian one. The title was not apparently used
by other Akkadian kings. It was used once by a Gutian king, and once by Utu-Hengal of
Uruk, who was overthrown by Ur-Nammu. Ur-Nammu himself did not use the title,
1 14 Lesson 10
presumably because of the limited size of his realm; note also that he did not use the divine
determinative. The title was used by all the other Ur III kings, and afterwards by various
later kings, in an Akkadian or Sumerian form.
- History
Ur-Nammu was killed on the battlefield, but no specifics of his death are known; the
literary work entitled "The Death of Ur-Nammu" is terse and unsure at this spot. Woolley
believed that Ur-Nammu and the other Ur III rulers (except the last) were buried at Ur, in a
building complex he referred to as the "Mausolea" of the Ur III rulers. Moorey has
recently questioned this:
The balance of available information, archaeological and textual ... suggests
that if the kings (and queen-mothers) of the IIIrd Dynasty of Ur were buried
in that city it was not in Woolley's "Mausolea" . ... Ur is not the only potential
site for these graves, for they might have been in a palace at Uruk, home of
the dynasty, or, less probably, even perhaps at Nippur ( 1 984: 1 8).
It was under the rule of Ur-Nammu's son and successor Shulgi that the Ur III empire
reached its greatest extent; Piotr Steinkeller calls him "the true builder of the Ur III state"
( 1 987b:20). There was a great deal of royal building; there was a reform in the calendar,
and much bureaucratic reorganization; Gadd says that "visibly under the impulse of the
king himself a most meticulous system of bookkeeping was instituted" ( 1 97 1 :617). Daniel
Snell also points to "a general economic stability during the middle years of the Ur III state,
a stability that seems a likely corollary of the middle kings' largely successful attempts to
maintain the empire their predecessors had bequeathed them" ( 1 982: 1 9 1).
Steinkeller ( 1987b:20-21 ) lists the following among the reforms of Shulgi:
1) the deification of S hulgi
2) the creation of a standing army
3) the reorganization of the system of temple households
4) the creation of a unified administrative system for southern and northern
Babylonia
5) the introduction of the bala taxation system, coupled with the creation of
a chain of redistribution centers ... which served to collect, to process,
and to distribute the state revenues
6) the creation of an enormous bureaucratic apparatus, as well as of a
system of scribal schools that provided highly uniform scribal and ad
ministrative training for the prospective members ofthe bureaucracy
7) the radical reform of the writing system
8) the introduction of new accounting and recording procedures and of new
types of archival records
9) the reorganization of the system of weights and measures
1 0) the introduction of a new calendar, the so-called Reichskalender, which
became the official caldendar throughout the Ur III state.
It was about half-way through Shulgi's rule when he began to conduct many military
raids. A number of these were directed towards the East, modern-day Iran. The details of
Lesson 10 1 15
these campaigns are rarely known to us, and in fact it is surprising how little historical
information we have about Shulgi, especially considering the fact that he ruled for almost
half a century.
Shulgi was also the subject of some thirty hymns, preserved to varying degrees. In
"Hymn B", he boasts: "I learned the art of the scribe from the tablets of Sumer and
Akkad"; he also refers to himself as "the scribe of Nisaba", the goddess of wisdom and
writing.
1 16 Lesson 10
Text l Oa
supplementary
Notes
This autograph is taken from Volume 1 of Sir Henry Rawlinson's The Cuneiform
Inscriptions of Western Asia, published in 1 8 6 1 . This volume was one of the earliest
collections of cuneiform inscriptions published. At that time, very little was known about
Sumerian. In fact, it seems that it was not until the year 1 869 that the word "Sumerian"
was attached to this language, by Jules Oppert.
It is very difficult to correctly copy texts written in a language which one does not
understand. This means that such autographs are sometimes slightly "off '. In this
particular autograph, some of the signs seem to differ from those in the other texts.
The shape of the ki-determinative in this line is quite different from the shape of the ki
sign in lines 9 and 1 1 . It is hard to say how much of this variation is due to the original, or
how much is due to Rawlinson.
8. There is no expected -ma at the end. This is most unusual, and one suspects that this
is an error of Rawlinson, rather than an error of the original scribe.
Lesson 1 1
This text was inscribed on a headdress or wig of diorite. D.J. Wiseman has pointed
out that this headdress was "scored on the underside to fit the rounded head of a statue".
He concludes that it was "intended to be fitted to an actual statue rather than be used
independently as an ex-voto object" (1960: 168 n.25).
munus woman
8 dabs to hold
n* am
Notes
Lamar The reading of the name in the first line is unclear, since the text is partially broken.
There is a commonly attested goddess whose name is usually written with the kal-sign (Le.,
the kalag-sign); it is variously read by Sumerologists as Lama2 (or Lama), Lamma2 (or
Lamma), Lamar, or Lammar. Since this word was apparently borrowed into Akkadian as
1 17
1 18 Lesson 1 1
lamassu, there was probably some kind of Ir/-Auslaut; Lieberman reconstructs the original
form as: ILamat/. (The word is further discussed below.)
Virtually everyone who has studied this text has read the first line as dLamar (re
gardless of the precise transliteration). However, there seems to be present a stroke of
another kal-sign, immediately after the divine determinative: d [Kall-kal. This may be a
different writing for the same goddess Lamar, or it may represent an altogether different
deity (a god dKal-Kal is elsewhere attested, but seems to be masculine). For simplicity
sake, and since only a single stroke remains of the problematic sign, the line will be
transliterated as dLamar.
Nin-gir-su Etymologically, "Lord of Girsu". nin is used here in the sense of en, "lord",
as in Text 1 .
Girsu was the sacred quarter of the city and state of Lagash (further discussed in
Lesson 22); it actually lay some distance outside the city proper. Ningirsu was the tutelary
divinity of the state of Lagash. His most famous temple was the E-ninnu (of uncertain
meaning, "House 50" ?).
Ningirsu seems to have been the local name for the god elsewhere worshipped as Nin
urta, a god originally of agriculture and storms, but also of war. The two were probably
independent deities who were very early identified with each other.
Nane This name is read by some Sumerologists as Nazi. She was a daughter of Enki,
and the goddess of Lagash. She was consulted for the interpretation of dreams. When
Gudea, the ruler of Lagash, had an odd dream in which a mysterious figure appeared, it
was Nanshe that he turned to for the explanation of the dream.
The cuneiform sign representing her name is the sign for the city of Sirara (one of the
places where she was especially worshipped), with an inscribed -sign. means
"fish"; the sign is in origin the picture of a fish. This and other evidence indicates that
Nanshe may originally have been some kind of fish-goddess.
Ba-ba6 The reading of the last sign is uncertain. The name is variously transliterated as:
Ba-y, Ba-, Ba-bU l l (and Ba-bu l 2 ) ' and Ba-w,\. She was the wife of Ningirsu, and
hence the city-goddess of Lagash. At times, Inanna herself is referred to as "Ba-".
zabar In older transliterations, each of the three individuals signs forming this compound
logogram is separately transliterated: ud-ka-bar. The etymology and writing are discussed
below.
zabar-dabS The etymology is discussed below. The function of this official is not too
clear. Jacobsen refers to him as "the official in charge of the bronze (table-wares, cups,
knives etc. of a large establishment, and possibly of the bronze weapons as well)"
( 1 957 :382 n.55).
The word was borrowed into Akkadian as zabardabbu. The CAD simply translates
the Akkadian term as: "an official". After a long discussion, it concludes with the remark
that this official was "(possibly), originally the weapon carrier of the king".
Lesson 1 1 1 19
bi-li The basic meaning of bi-g appears to be something like "chann" or "attraction". (The
Akkadian equivalent, kuzbu. is glossed by the CAD as: "luxuriousness, abundance,
attractiveness, charm, sexual vigor".) It also has the derived meaning "headdress" or
"wig". The gudug-priests discussed in Lesson 1 9 are attested as wearing a bi-g.
munus Also transliterated as mi . Particularly in older works, it is transliterated as sal.
dabS In older transliterations, dib (and, incorrectly, dib4).
dim While du is used for the (re)construction of more solid objects, such as palaces,
temples, etc., dim is nonnally used of smaller, hand-made objects.
The Akkadian equivalents for du and dim are not neat. dim is nonnally equated with
'
banu, but du is equated both with banu and epeu.
Am This sign is composed of two elements: Tt which nonnally has a syllabic reading ,
'
and * which nonnally has a syllabic reading an. It is not clear how these two signs
'
Text 1 1
4ClFr ET
rrIJ1I
r1! .: ;:'
. . .
*:'
O
+-f r4 k'=l
Lesson 1 1 121
[
appositives. The first is the noun-adjective combination, nitab kalag-gf!.. The second is the
genitive phrase "king of Ur". By itself, this last genitive phrase would be expressed by:
J
lugal.Urim.ak. This nominal phrase may be diagrammed as:
[nam.til] S UIgi .ak.e
nitabkalaga
lugal.Urim.ak
This results in a succession of two genitive markers, followed by the marker for the
tenninative case: .ak.ak.M. In the script, this is reflected as: ... Urimskt ma-ka-e.
Since the Ikl of the first genitive marker .ak is followed by a vowel, it is pronounced
and written in the script. The Ikl of the second genitive marker .ak, however, is syllable
final before a consonant. In such cases, the Ikl does not show up in the script. The
problem of the amissable word-final consonants was discussed earlier. To some degree,
that same problem is present when these consonants are syllable-final, not just word-final.
That it, it is not sure if these consonants were pronounced or not: it is not known if this is
an orthographic or phonological problem. The sequence .ak.ak.e, for instance, is common,
but the Ikl of the second genitive marker does not appear to ever be written in any of these
Occurrences. In general, it seems that the Ikl of the genitive marker does not appear in the
script, when it is in syllable-final position followed by a consonant. It is transcribed here
within parentheses.
The ka-sign includes the Ikl of the first genitive marker, and the lal of the second
genitive marker. This is yet another instance where the script does not follow the
morphology, but rather approaches the syllabic structure of the spoken language. This
122 Lesson 1 1
first genitive marker is for [Nin] . [Girsu] .k; the second is for [Ur] . [Ningirsu.k] .ak.
However, as discussed in Lesson 1 concerning the name Ur-Nammu, there is some
indication that the genitive marker in PNs was lost, and so it is not indicated in
transcription.
Lines 7-10 form the ergative nominal phrase, expressing the agent of the transitive
verb. The ergative case-marker .e appears at the end of line 10. The nominal phrase is
complicated by the presence of all the appositives. Line 7 is a PN, Babaninam. Lines 8-9
ru;e an appositive, describing Babaninam as the "zabardab of Ur-Ningirsu". This would be
expressed as: [zabardab]. [Urningirsu] .k, but Ur-Ningirsu himself is described as "the
beloved lord of Nanshe", an,.. appositive. "Beloved lord of Nanshe" is: [en.ki.aga.a] .
[Nane].k, a more complicated genitive phrase than any seen up to now: ki.aga.a modifies
en, and the combination is itself the first element of the genitive phrase. Thus, the nominal
chain describing Ur-Ningirsu is: Urningirsu, en.ki.aga.a.Nane.k. All of this is the second
element of a genitive phrase, with zabardab being the first element: [zabardab] . [Urningirsu,
en.ki.aga.a.Nane.k] .ak, all of which is an appositive to Babaninam. This may be dia-
]
gr
1;:: b
am
rUrningirsu ]
.e
l
amm .ak
Len.ki.aga.a.Nane.k
The writing of the end of the nominal phrase is as expected: ... dNane-ka-e4.
1 1 . munus is "woman", nam-munus is an abstract, "womanhood".
This line must mean something like "her beauty of womanhood", that is, "her woman's
beauty". If so, then it is a genitive phrase, followed by a possessive-suffix: [bili.nam.
munus.ak] .ani.
The basic rule for the genitive marker as presented up to now has been: /ak/ after
consonants, /k/ after vowels. Here, however, there occurs: nam-munus-ka-ni; the vowel
/a/ of the genitive marker /ak/ does not seem to appear in the writing.
Such writings - where the /a/ of the genitive marker does not appear after a consonant
- are not uncommon; the next example is in a formulaic phrase appearing in Text 13. It is
difficult to say whether such writings tell us something about Sumerian orthography, or
Sumerian phonology, or Sumerian morphology; several interpretations are possible. In
order to make the written form fit our understanding of the grammar, one school of thought
would read the first sign as munusa, instead of munus. This would produce munusa-ka-ni,
accurately reflecting munus.ak.ani. Parallel phenomena occur outside of the Ur III corpus.
For example, in the inscriptions of Gudea - inscriptions highly localized to one time and
place - "his king" is expressed by both lugal-ni and lugal-.!!-ni. In order to make lugal-ni fit
more accurately our understanding of Sumerian, some scholars would read the two signs as
lugala-ni, and not lugal-ni.
This school of thought was particularly adumbrated by Adam Falkenstein, who saw
similar phenomena elsewhere in Sumerian grammar. He coined the term "Uberh!tngende
Vokale" (in English, "overhanging" or "overlapping" vowels) to describe just such
writings. This school would thus see such writings as an orthographic problem. (The
standard sign-lists, however do not seem to recognize a reading *munusa for the sign in
124 Lesson 1 1
question.)
A different view sees this as a morphological (or phonological) problem. Mamoru
Yoshikawa feels that the genitive marker was sometimes Ikl after a consonant, not only
after a vowel. However, he cannot posit rules for the distribution of lakl and Ikl after
consonants. Yoshikawa also believes that there are cases where the genitive marker is
I ak/, not Ik/, after a vowel.
This is not a problem which can be solved here. But it must always be kept in mind
that the writing system of Sumerian never accurately reflected the spoken language. It is
possible that although a scribe spoke /lugalanil, he was perfectly happy to write l!!gal-ni,
because with just these two cuneiform characters, he knew what to read. Why bother to
write an i!-sign if the context makes the presence of a spoken lal obvious?
This entire line is the direct object (patient) of the verb dim, and therefore is in the
absolute case.
Discussion: structure
It is difficult to see the basic structure of this text, because of the presence of so many
appositional phrases. Its substance, however, is: "Babaninam fashioned a wig for Lamar,
for the sake of the long life of Shulgi":
- Writing system
The innocuous-looking word zabar illustrates some of the intricacies of the Sumerian
writing system. There are no metals native to Sumer; rather, all had to be imported. Thus,
zabar i s not a native Sumerian word; it was borrowed from some unknown language.
Hallo says "in general, it may be supposed that the basic metal names are non-Sumerian
'Kulturworter' or 'Wanderworter' which were adopted together with their referents"
(1963: 140). In Akkadian, the word for bronze is siparru. Akkadian may have borrowed
this word independently from the same language that Sumerian borrowed it from, or, much
more likely, borrowed it directly from Sumerian. In either case, the form siparru is a little
odd; it would seem to derive from */sipar/, not Izabar/.
One way to solve this discrepancy is to assume that in earlier Sumerian the word for
"bronze" was, in fact, Isipar/, and that Izabarl represents an inner-Sumerian development.
The change of Ipl > Ibl is not surprising; voicing of inter-vocalic voiceless consonants
happens in many languages. The change of initial Isl > IzI is less easily explained, but
there are other parallels to this change in Sumerian. The difference in vocalization between
the two forms is more interesting. As will be discussed in Lesson 20, there is a fair amount
Lesson 1 1 1 25
of evidence to show that Sumerian has undergone a rather wide process of vocalic
assimilation. In words originally containing two vowels of differing quality, one vowel has
assimilated to the quality of the other. In our particular case, an original */i-al has become
la-al.
Thus, Izabarl can be derived from Isipar/, using sound-changes which are elsewhere
attested in Sumerian. (Unfortunately, not enough is known to date these sound changes.)
This then would represent a case where the Akkadian word has actually preserved a more
archaic form of the word than has Sumerian. Presumably Akkadian borrowed it from
Sumerian before these changes took place. Similar instances will be discussed in Lesson
20.
The pronunciation of this word in historic Sumerian as Izabar/, and in Akkadian as
Isiparru/, is known from lexical lists, where these words are spelled out as za-bar and si
par-ru. A next question is, how does the pronunciation Izabarl "derive" from the three
signs ud-ka-bar? In the word nidba, for instance, there was no obvious way to phonetically
relate the pronunciations Inindabal or Inidbal to any pronunciation of the individual signs;
that is, the word was more than the sum of its parts.
Since one of the three signs forming the word for "bronze" is the bar-sign, it seems
reasonable to assume that Izabarl derives phonetically from these three signs. B ut how?
One possibility might be to read ud-ka as zax . This type of approach is favored by many
Sumerologists, who try to make the writing system better fit Sumerian pronunciation.
However, there does not seem to be any other, independent, evidence which would justify
positing a reading zax for this sign, and the standard sign-lists do not recognize such a
value.
However, a further complication must be introduced. In the earliest Sumerian, the
word for "bronze" is not, in fact, written ud-ka-bar. Rather, it is regularly written as
KAxUD-bar, that is, with a ka-sign containing an inscribed ud-sign, followed by the bar
sign. For example, in a royal inscription of king Uruinimgina of Lagash, the word for
bronze appears as: L .
It is difficult to explain such a writing. Some scholars have posited a reading for
KAxUD; this is accepted by some sign-lists, although with reservation. However, Fal
ken stein has pointed out evidence that KAxUD can be read as sil 9 . This reading is
accepted by the standard sign-lists. Perhaps, then, the word should be transliterated as:
si 19-bar. Armas Salonen, in fact, transliterates this word as si l9 -bar for the older period,
but as zabar for the "nachsumerische" period ( 1961 : 108). In this interpretation, which is
probably correct, si l9 -bar represents an older pronunciation of the word. B ut the same
cuneiform signs continued to be used, when in spoken Sumerian the word had changed to
Izabar/.
- Writing system
this is true to some degree even in Old Babylonian texts. This leaves the origin of the nin
sign up in the air.
- Overhanging vowels
The problem of the overhanging vowels has generated a good deal of polemics in the
course of Sumerological studies (the latest discussion is Yoshikawa 1980). This over
hanging vowel seems to be mostly la/, although individual cases of overhanging le/, liI,
and lul have been posited. Probably most Sumerologists accept their existence, but there
are dissenting opinions.
In Falkenstein's view, the readings in lal represent older forms of S umerian words.
That is, at one time these words were pronounced with a final la/. The word for "king",
for example, was originally *Ilugala/. At some (prehistoric) point, these final lals were
dropped; the word for "king" became Ilugal/, but the sign used to represent this word
could be used for either the newer value Ilugal/ or for the older value *Ilugala/. Since
both values co-occurred, a scribe could write "his king" as either lY.gal--ni, or lY.gala-ni,
both representing /lugalanil. (Presumably, the prehistoric form would have been *Ilugala
anil or *Ilugala-nil.)
Other Sumerologists question their existence; they see varying phonetic factors at
work. It was mentioned above, for example, that Yoshikawa believes that the genitive in
Ikl occasionally occurred after consonants, not just vowels; thus, "the son of the king"
might have been pronounced Idumulugalk/. However, Yoshikawa could not state any
general rules for the distribution of Ikl and lakl after consonants. For other overhanging
vowels (e.g., in the verbal system), he has other explanations.
In writings of the lY.gal-ni type, it has been posited that the lal was dropped:
*Ilugalanil ) Ilugalnil (this was Poebel's position). Without going into details, it can be
seen that such an explanation raises more questions than it answers, such as the co
occurrence of writings like lugill-ni and lugill -fl-ni in the same time and place.
Is there any independent evidence which justifies the view that certain signs contain an
overhanging vowel? Here the evidence varies, and is difficult to interpret. The fact that the
nominalized forms of the verb "to love" are written both ki-5!g5!2 and ki-g-g might seem to
indicate that this one sign can be read as lagl or laga/. However, it is also possible that
a
g is here a phonetic complement, and the form should be understood as: ki-5!g g ; or it
might be that these writings reflect phonological problems of particular roots ending in a
vowel.
The Akkadian lexical tradition is likewise ambiguous. Lexical lists do provide
readings with lal for some signs, but for the most part they do not (for example, they show
no evidence of a reading lY.gala). And, some of these readings with an overhanging vowel
may very well result from the Akkadian scribes encountering the same problems in the
writing system that we feel. These scribes may have anticipated some modem
Sumerologists, by generating readings in la/, in order to make the writing system more
closely fit the pronunciation.
Part of the problem may result from a misunderstanding of the nature of the Sumerian
writing system. Because the expression "malt house" is written -bappir, for example, and
Lesson 1 1 127
not written -bappir-ra (for the assumed e.bappir.a(k , Falkenstein would say that the
second sign should be read as bappira. But it is easier to say that the Sumerian scribe felt
no need to write any indication of the genitive marker; such scribes were content to write
bappir, even if they pronounced it /ebappira(k)/. Falkenstein's school is an attempt to
make the writing system more closely resemble a transcription of speech, and this is not
how the writing system should be understood.
It is true, however, that there are other problems to be resolved. For example, "in the
land of Sumer" is normally written kalam-ma in the Gudea texts, for kalam.a. But once,
apparently, this locative phrase is written : kalam. Falkenstein would read this kalama.
Similarly, "on the tablet", written just dub, would be read by Falkenstein as duba. These
writings raise questions, but they may simply reflect an earlier period in Sumerian
orthography, when it was not necessary in general to write case-endings.
- Loan words
As just discussed, "bronze" is zabar in S umerian, siparru in Akkadian. The word for
"copper" is urudu in Sumerian, weru in Akkadian; they are usually spelled out in lexical
lists as !i-ru-du and -ru-.Y.. The ultimate origin of the word(s) is unknown. Both urudu
and wen1 may reflect one pre-Sumerian substrate word; they have even been connected
with the Indo-European word which ultimately appears in English as the adjective "red".
Curiously, in late Akkadian the word for "copper" also appears as urudG . Eduard Kutscher
says that "this 'Akkadian' word was artificially coined by Sennacherib's scribes (and used
only by them) from Sumerian urudu (= Akkadian wen1) 'copper'" ( 1982:225).
- History
- History
Other wigs have been found at various sites in Mesopotamia, although none of them
bear an inscription. For example, a wig of steatite was found at Uruk. Only 2.5 cm. long,
it was apparently designed to fit a statue (the dating is uncertain; before Nabonidus). More
recently, somewhat similar wigs have been found at Ebla.
128 Lesson 11
Text I I a
supplementary
Lesson 1 2
This text was inscribed on the foot of a vase of marble. No photograph is available.
(; ud U4) day
mu-sar-ra inscription
u ... ur to erase
stv:=r bi
ib
-q b
f3=-:J da
Notes
dBU-ga-me This is now the preferred reading (or at least the more original reading) of
the name more familiarly known as "Gilgamesh". The apparent meaning of the name is:
"the old man (bi1-g is (now) a young man (mM)". It is not known when the change of
initial /b/ ) / g/ took place; an Old Babylonian omen text has the spelling dGe-el-g.
Some read the last sign of the name as mes.
mu-sar-ra sar is "to write". sar-ra sar.a, the passive participle: "something written".
=
mu has many meanings (e.g., "name"); mu.sar.a. is something like "a written text", or
"inscription". It was borrowed into Akkadian as musarU, translated by the CAD as: " I .
object bearing an inscription, 2. inscription".
u . .. ur This is a compound verb. u means "hand", and ur means something like: "to
move or drag (something)". u is thus the historical patient of ur: "to move the hand over",
i.e., "to erase". The verb takes its complement in the locative case. (The Akkadian
1 29
130 Lesson 12
equivalent is pasatg, glossed by AHw as: "tilgen, ausloschen".) ur also enters into the
formation of other compounds, for example: gis ... ur, "to harrow" (literally, "to drag wood
(over the ground)".
Some believe that the root ended in lul, and so it i s also transliterated as su ... uru 12'
nam ... kurS nam is apparently the (historic) direct object (patient) of kur5 ; the meaning is
something like "to cut a decision (against)". It usually takes its complement in the
comitative case. The word is further discussed below.
Lesson 12 131
Text 1 2
132 Lesson 12
marker in .a.
The most common relative marker is the noun ud (literally, "day"); others also occur.
Occ asionally the relative marker is deleted.
The essence of the relative clause is: "He built the temple of Nanna". This
independent sentence would be: e.Nanna.(k).0 mu.(n.).du.0. To form the relative clause,
this independent sentence is nominalized in .a. It is then placed in apposition to the relative
marker ud: ud [e.Nanna.(k).0 mu.(n.)du.0].a.
It is instructive to compare this relative clause with the relative clause in Text 7. In that
text, "the one who built the temple of Enlil" was expressed as:
lu [e.Enli1.a(k).0 i.n.du.0].a.
The construction here is much the same, but the relative marker is ud instead of lu:
ud [e.Nanna.(k).0 mu.(n.)du.0].a
Finally, the entire complex is put into the locative case, marked by .a (the second .a in
the transcription of line 9).
The function of the locative in .a here is "on": "on the day that". It is also possible to
find other case-markers, such as the ablative case-marker ta ("from, since"), or the
terminative case-marker Se ("until").
A literal translation of this entire clause would be: "on the day that he built"; this
captures both the force of the Sumerian locative, and the force of the relative clause. In
more idiomatic English, however, one may say "at the time when", or simply "when":
"when he built the temple of Nanna, he made a votive offering".
In this temporal clause, the verbal chain uses the conjugation-prefix mu, and the
personal-affix .n does not appear in the script. In the relative clause in Text 7, the
conjugation-prefix was 1. and the personal-affix .n appeared in the script. It is not easy to
understand the reasons for such alternations.
Although the transcription indicates both the .a of the nominalizer and the .a of the
locative case-marker, it is reasonable to assume that some kind of vocalic contraction took
place; one never finds two /als in the script in such a construction.
Lines 8-9 form the subordinate clause; lines 10- 1 1 form the main clause. In Sumerian,
the subordinate clause regularly precedes the main clause.
1 2- 1 5 . These lines express a curse; essentially the same wording occurs in other Su
merian inscriptions.
Lines 12- 1 3 are a relative clause, serving as the complement of the verb in line 15. Its
meaning is: "the one who erases this inscription". The relative marker is lu; the relative
clause is marked by the at the end of line 13.
Su ... ur is a compound verb. bi is another conjugation-prefix, in addition to the mu and
1 already seen. It is almost always written with the bi-sign. It will be further discussed
below.
All the verb forms seen up to this point have been in the gamn! aspect, used to express
action in the past. The verb form in this line, however, is being used to express a future
value: "whoever will erase", or "whoever might erase". Therefore, it is put in the man1
aspect.
It is common to speak of the "gamtQ-root" and the "marfi-root". The man1-root is
1 34 Lesson 12
fonned from the bamt!!-root in several different ways. The fonnation used for any par
ticular verb is lexical; that is, it is not predictable. It is not clear exactly how many classes
of maru fonnation exist. Yoshikawa has established three different classes:
(1) "Reduplication" class. The man1-root is fonned by (graphically) reduplicating the
bamt!!-root. Thus, "to return": bamt!!, g4; maru, g4-g4.
Roots of the pattern eve seem always to lose their final consonant when reduplicated.
Thus, "to place": bamt!!, gru:; man1, .M.-.M (always written with the g!!-sign).
(2) "Alternation" or "replacement" class. An entirely different root is used for the
maru. This root is non-predictable from the bamm-root. Thus, "to speak": bamm, .d!!g4 ;
maru, . dug4 and are two entirely different signs. However, there are cases where a
bamt!!-root and a maru-root will be written with the same sign. Thus, "to go": bamt!!, gin;
maru, duo The gin-sign and the du-sign are the same! In such cases, it is only the gramma
tical context which indicates whether the sign is to be read as the bamt!!-root or as the
maru-root.
(3) "Affixation" class. This is fonned by addition of .e to the bamt!!-root. This .e is
variously referred to as the "maru-element", the "maru-affix", or the "man1-suffix". Thus,
"to build": bamt!!, du; maru, du-. This class is the most common fonnation of the man1.
It is the fonnation used with the verb in Text 1 2, ur: ur.e. Here, however, the maru-suffix
has contracted into the nominalizer .a, and so it does not show up in the script.
In general, the particular maru fonnation for any specific verb is not always known.
And, some roots fall into two (or even all three) classes. As mentioned above, ur is a
member of the affixation class, but reduplicated man1 fonns also seem to occur. In later
Sumerian, combinations of these classes sometimes occur; e.g., a reduplicated root
followed by the maru-suffix. These cases have not all been explained.
The use of the maru, instead of the bamt!!, entails rather complex changes in Sumerian
morphology, particularly in the distribution of the personal-affixes. For a transitive verb:
in the bamm, the personal-affix slot before the verbal root cross-references the agent, and
the personal-affix slot after the verbal root cross-references the patient. In the maru,
however, it is j ust the opposite. The personal-affix slot before the verbal root cross
references the patient, and the personal-affix slot after the verbal root cross-references the
agent. For example, "The king built the house" is:
(1) lugal.e e. mu. .du.
But, "The king will build the house" is:
.s: *'
(2) lugal.e e.0 Lb.du.e.0
In (2), the ergative case marked in .e is cross-referenced by the .0 after the maru-
suffix. The absolute case marked in .0 is cross-referenced by the .b before the verbal root.
.b is used here to cross-reference inanimate antecedents; .n is used to cross-reference
animate antecedents.
That is, the case-markings on the nominal participants in the sentence are the same in
both aspects: lugal.e and e.0. However, the use of the personal-affixes is quite different.
Thus, in the verb fonn in this line, u.bi.b.ur.e.0.a.(d), the .e is the maru-suffix, and
the .0 cross-references the third-person agent. (This analysis of .e.0 is not universally
accepted, and a different view will be mentioned below.) The .b. before the root cross-
Lesson 12 135
u bi . b . ur . e . 0 . a
(1) (2) (3) (4) (5) (6) (7)
(1) nominal element o f compound verb
(2) conjugation-prefix
(3 ) personal-affix cross-referencing patient )
(4) verbal element of compound verb
(5) manl-suffix
(6) personal-affix cross-referencing agent (unexpressed)
(7) nominalizer, forming relative clause
with proper names; the name was felt as a unit, and the grammatical case-endings added
directly to the complete unit.
1 5 . b is a form of the "desiderative" or "precative" modal-prefix. As discussed in Lesson
1 , the very first element in the Sumerian verbal chain is an optional modal-prefix; there are
several of these.
The desiderative mood expresses wishes and indirect commands in the third person:
"Let him/them, may helthey", etc. The basic form of the desiderative is Ibel (written b);
before the conjugation-prefix ba, it regularly becomes Ibal (written b).
One of the more difficult questions in Sumerian morphology is the nature of the root
used after particular modal-prefixes. Some modal-prefixes use the bamm-root, others use
the man1-root, and still others use both, under conditions which are not always clear. A
second problem is the use of the personal-affixes. In some moods, the pre-verbal root slot
cross-references the agent. In others, it cross-references the patient. (This differentiation is
apparently irrespective of whether the bamt!!-root or the man1-root is used.)
With transitive verbs, M. is regularly construed with the mam-root; with intransitive
verbs, it is construed with the bamm-root. kurS is a member of the affixation class, and so
the form is: kurS -' The writing is morphemic in this line; in other cases the final consonant
of the verbal root may be graphically reduplicated (e.g., kurS -re).
In the desiderative, the agent is cross-referenced by .0 after the maru-suffix.
As with all compound verbs, the first element of the compound (here, nam) precedes
the entire verbal chain, including the modal-prefix.
ba is a conjugation-prefix not seen previously. It is usually assumed that it is related to
the conjugation-prefix bi , an example of which occurred in line 13. It is further discussed
below.
da is the dimensional-prefix which cross-references the comitative case. Here it cross
references the comitative case marked by the presumed .da at the end of line 13. The
original meaning of this compound verb may have been something like "to cut a decision
against".
To summarize the verb phrase, nam . . . kurS is a compound verb, with nam the (his
toric) patient. b is a form of the modal-prefix for the desiderative; with transitive roots, it
uses the maru form of the root. ba is a conjugation-prefix. da is a dimensional-prefix,
cross-referencing the comitative. The verb form may be diagrammed as:
nam be ba da kur . e 0
(1) (2) (3) (4) (5) (6) (7)
(1) nominal element of compond verb
(2) modal-prefix
(3) conjugation-prefix
(4) comitati ve dimensional-prefix
(5) verbal element of compound verb
(6) mam-suffix
(7) personal-affix cross-referencing agent (Gilgamesh).
Lesson 12 137
Discussion: structure
The structure of this text is:
- Phonology
The second element of the compound verb nam... kur5 is variously transliterated as
kU5 ' kud, and kuru 5 (the latter with an overhanging vowel). The Idl - Irl alternation helps
show that the (amissable) Auslaut of this root was probably some kind of Irl sound which
did not exist in Akkadian. The Akkadian writing system sometimes reproduced it by Irl,
and sometimes by Id/. Modern Sumerologists sometimes transliterate it as dr, cIf, or t. It
seems to occur as the last consonant in some dozen or so verbal roots; another instance
occurs in Text 17. Its presence as the first consonant or medial consonant of a root is much
harder to detect.
The kur5 -sign can also be read tar. Confusingly enough, there appear to be two
different verbs: nam-kur5 meaning "to curse", and nam-tar, "to decide the fate ofIfor", "to
'
decree a destiny for". The boundary between the two expressions is, however, sometimes
unclear, and occasionally transliterations are less than precise in differentiating between the
two; both expressions are sometimes found transliterated as nam-tar.
- Moods
The morphology of the moods in Sumerian is quite complex. The single most
important work to unravel them is Dietz Otto Edzard 197 1ff; this is a series of articles
which it pays to keep close at hand. Some of Edzard' s conclusions were modified by
Burkhart Kienast ( 198 1 b).
- Conjugation-prefixes
In the paradigms presented in this book, model verbs in the bamt!! are generally cited
with the conjugation-prefix mu, and those in the maru are generally cited with the con
jugation-prefix 1. While this does represent the most common distribution (at least in the
Ur III royal inscriptions), it is also possible to find verbs in the bamt!! with the con
jugation-prefix 1 and verbs in the maru with the conjugation-prefix mu (although this latter
is rather rare).
It is sometimes stated that the conjugation-prefix ba "represents" bi with an additional
locative marker of some kind. It is difficult to say whether or not such a statement is
138 Lesson 12
- Comitative - patient
In the analysis given here, lines 12- 1 3 are in the comitative case. However, the
apparent absence of the case-marker da does give one pause. Some people believe that
lines 12- 1 3 are the direct object of nam ...kurS . The comitative dimensional-prefix would
represent what Poebel called the "erstarrter Gebrauch des Infixes", that is, "frozen use of
the [dimensional] infix". In Gragg's study of the dimensional-prefixes in Sumerian literary
texts, he found a large number of instances where a comitative dimensional-prefix occurred,
without any corresponding comitative case-relationship.
This is not impossible. However, there is a more general issue here. In the compound
verb nam...kurS nam is apparently a (historic) direct object (patient) of kurS . Now, it is
'
usually believed that Sumerian does not permit two patients in one sentence. In cases
where one might expect two patients, one of them will be expressed through one of the
adverbial cases. In the immediately preceding lines, for instance, the English translation
was "to erase this inscription". The direct object (patient) in Sumerian was u; the direct
object in English, "this inscription", was expressed via a locative: "to move one's hand
over". If it is true that Sumerian does not tolerate two patients, and if it is also true that this
rule applies to historic direct objects (patients) of compound verbs, then lines 12-13 must
be marked by another case. Given the presence of the dimensional-prefix da in the verbal
chain in line 15, this would most likely be the comitative case.
The argument that Sumerian does not permit more than one patient in a sentence is
based on general linguistic theory, and on empirical observations in Sumerian. Some
linguists would say that no language has more than one patient (at least in the deep
structure); if two seemingly occur, one must be in an adverbial relationship. However, it is
not clear if such a constraint would apply to compound verbs. Even though the first
element of many compound verbs is historically the patient of the verb, it is not always sure
if it functioned as such in historic Sumerian.
Lesson 12 139
-Conjugation
Following is the paradigm for the mam of the transitive verb in the singular. The
model used is sar, which is a member of the affixation class. The conjugation-prefix used
here is 1.
first person singular 1-sar-re-en Lsar.e.en
second i-sar-re-en Lsar.e.en
third i-sar-re Lsar.e.0
The first and second persons singular are identical in form. The final 1nl often does
not show up in the writing.
If the root ends in a vowel, there is frequent assimilation of the Iel of the marO-suffix
into the vowel of the root.
Some Sumerologists analyze the morphology of these endings as: .e.n, .e.n, .e.0.
That is, the first and second person markers are .n, not .en. This is a thorny issue, which
cannot be resolved here.
For verbs of the reduplicating class (the model is gar, "to place"), the forms are:
first person singular 1-R-g-en Lgaga.en
second l-g-g-en Lgaga.en
third i-g-g Lgaga.0
This interpretation of the maru of verbs of the reduplicating class is essentially that of
Yoshikawa. However, all such reduplicated forms end in a vowel, which is subject to
contraction with the lel of the ending .en. This means that in texts, such forms as the
following are encountered: i-M-M--en, i-gfi-gfi-an, i-gfi-M-, etc. Other Sumerologists
have argued that such writings indicate that Yoshikawa's analysis is incorrect.
The marker .0 in such marO forms as Lsar.e.0 has been treated here in two ways.
First, it was called a personal-affix, cross-referencing the transitive subject. Second, it was
called a marker for the third person. As discussed in Lesson 1, these are not contradictory
interpretations, but are rather two ways of saying the same thing.
- marO formation
It was Yoshikawa who established the three clases of mam formation discussed
above. Edzard has expanded this into five classes:
(1) "Unchanging": The hamtu-root and the maru-root are the same. This cor-
responds to Yoshikawa's "affixation" class; Edzard does not believe that .e is a marO
marker.
(2) "Reduplicating": Same as Yoshikawa.
(3) "Root-varying": The two roots are different, but similar phonetically: "To
approach" is te in the bamn!, but teg in the mam and teg are the same sign).
(4) "Replacement": Same as Yoshikawa.
(5) "Irregular": These do not seem to fit nicely into the other four categories.
140 Lesson 12
example, .lli4 -ur4 ' "to look for and gather up", almost always appears this way.
Presumably, this is because it is almost always used with a plurality of objects. Such a
lexicalization has a parallel in the Semitic languages, where certain roots are lexicalized in
certain stems. For example, the Akkadian buu, "to look for", only occurs in the D-stem.
In context, it is often very difficult with certain verbs to decide whether marii re
duplication or free reduplication is present. And if it is a case of free reduplication, it is
often not easy to see its function.
It is probable that reduplicated forms (of whatever kind) were phonetically reduced in
speech. There is evidence for this from texts written in syllabic orthography (Appendix 2)
and from unusual occurrences of syllabic writings within normal orthography. For
example, an expected biz-biz- appears once in Gudea as: bi-bi-ze. Few details of such
phonetic reduction are understood.
- Origin of cases
It has been speculated that the Sumerian case-endings (some of them, at any rate) were
originally nouns. In particular, it has been claimed that the comitative case-marker da is in
origin the same da meaning "side", seen in the expression "king of the four quarters",
lugal-an-ub-da-limmuba. This is not impossible, but it is harder to find an etymology for
the other case-markings.
- Research in Sumerian
It is sometimes fairly easy to understand the meaning of a Sumerian text, and even
relatively easy to describe, on surface terms, what we see, but it is much more difficult to
understand exactly what is happening. For example, based on context and on parallels in
Akkadian, Phoenician, and Aramaic texts, all Sumerologists would understand the last lines
of this text to mean, "May Gilgamesh curse the man who erases this inscription",
regardless of the presence or absence of a comitative case-marker, regardless of the
presence or absence of a dimensional-prefix, regardless of the distribution of the personal
affixes, etc. However, unless such details are well understood, it is much more difficult to
figure out the meaning of really complicated passages.
This text also illustrates the problems encountered in doing research in Sumerian. An
obvious question which arose when discussing the presence or absence of a comitative .da
at the end of line 13 was, "How is this verb construed in its other occurrences? Does it use
the comitative or some other construction?" Unfortunately, without an up-to-date Sumerian
dictionary, such questions are not easy to answer. One can look at the existing dictionaries,
or texts with glossaries, but without a painstaking examination of many sources, it is
impossible to be sure that all instances of any particular word have been found.
-Curse-formulae
It is not uncommon for votive inscriptions to be provided with a curse-formula.
Typically, the first part of an inscription will form a straight-forward text; the curse is
tacked-on at the end. In the Ur III texts, only a limited number of curse-formulas occur; the
next occurrence is in Text 15.
Lesson 12 143
- History
The historical Gilgamesh was the fifth king of the First Dynasty of Uruk, which falls
within the Early Dynastic IT period (about 2700-2500 BC). No inscriptions of his are
preserved, or contemporary references to him, but there are a few inscriptions of his
approximate contemporaries.
The first attestation of him is in a god-list from Fara, where his name is written: dBiI
PAP-g-mM. It is difficult to say exactly how these signs represent the name /BilgameS/;
his name is spelled several different ways in the course of Mesopotamian tradition.
Gilgamesh seems to have been very popular with the Ur III kings; according to Jeffrey
Tigay, "The kings of Ur ITI regarded Gilgamesh as something like their personal god"
(1982: 1 3 n.50). Michalowski speculates that the "Gilgamesh stories were made part of the
school curriculum during the Ur III period" (1987:66). In his hymns, Shulgi refers to
Ninsun, the mother of Gilgamesh, as his own mother, and he refers to Gilgamesh as his
brother.
144 Lesson 12
Text 1 2a
supplemen tary
r r
1T Tt:>
i * ZZ
m <T : :}
I
5
z t: r:J
l' f
r m
2ffi -<r EU IUIl
m:A(/J t=4;> *
f tr>
*-7 t:>
2
m tlti
W f < ma tt> t
10
H H
Lesson 12 145
Note.
-dur-an-ki The name of the ziggurat at Nippur. Etymologically, "the temple, the bond
of (= between) heaven and eanh", E dur.[an.kiJ.k.
Text 12a was found outside of a controlled archaeological COnlext. The following
photograph is of a door socket bearing another copy of the same inscription, found in situ:
A. door socket. this granite block bore a door post in its cavity. With the discovery o f the thick buttressed
wall we had only the outside of a monumental building. To get inside. two plckmen dug a five.yard square
shaft beside the inner wall face. Walls and floors were soon located, but the room was larger than the shaft.
TIle few objects found did not identify the building. To learn more about the area with a minimum of
effort, we tunneted atong one wall for twenty feet, then another (or thirty feet to a corner where there was
a doorway. We dug through this, first finding a disappointingly uninscribed door socket, then. in a brick
box tower down, another with cuneiform wedges. The writing is that of king Shulgi, second ruler of the
Third Dynasty of Ur. commemorating his rebuilding of the temple of Inanna. With this, we had the cult
spot of Inanna at Nippur.
Lesson 1 3
This text is a standard inscription of Amar-Sin, the son and successor of Shulgi; he
ruled from 2046 to 2038 BC. The text exists in many copies; Text 13b is a stamped brick.
Sign-list and vocabulary
mu ...pad (w to propose
Notes
Zuen This seems to be another name of Nanna, although it is not clear why he had two
names. Jacobsen thinks that the term Nanna refers specifically to the god's role as the "full
moon", and Zuen refers to his role as the "crescent moon". It has also been suggested that
Zuen is the Akkadian equivalent of Sumerian Nanna; that is, they are two different names
for the same deity. However, there is no obvious Semitic etymology for Zuen.
The Akkadian equivalent of this DN is usually transcribed as either "Sin" or "Suen".
The Sumerian word was also borrowed into Akkadian as a common noun, appearing as
sInu, suenu, sinnu, and innu. It is glossed by the CAD as: " 1 ) the moon 2) crescent
shaped or semi-circular object".
The writing is discussed below.
Amar-dZuen Etymologically, "young bull of Zuen", amar.Zuen.(ak). The name of this
ruler is often transcribed "Amar-Sin" or "Amar-Suen", which are really Akkadianized
transcriptions.
Early scholars believed that this PN was Akkadian. The Akkadian equivalent of amar
is bOru. Therefore, the name appears in some older secondary literature as: "BOr-Sin", or
something similar. Almost all modern scholars believe that the name is Sumerian, although
there is really not much evidence to prove this.
147
148 Lesson 13
Nibru One of the more ancient and important cities in Mesopotamia; Kramer has called it
the "spiritual and intellectual center of Mesopotamia". Jacobsen has said:
From the very beginning of historical times Nippur and Enlil were recog
nized as an undisputed source of rule over Sumer as a whole, and kings of
Sumer would derive their authority from recognition in Nippur rather than
from their own city and its city-god ( 1957:139).
The modern name of the site is Nuffar. It was the first tell to be excavated by
American archaeologists (the University of Pennsylvania, beginning in 1 887). Over thirty
thousand tablets were found, mostly in Sumerian, ranging from the third to the first
millennium BC. Large numbers of these texts are still unpublished. The vast majority of
our Sumerian literary texts are Old Babylonian copies found at Nippur.
The etymology and writing of the name are discussed below.
sag-us The exact etymology and meaning are not sure. , as seen previously, means
"head". us has several meanings. It often means "to lie against", "to lean against". (It is
used as a logogram to represent Akkadian emedu, "to lean against, to reach".) It may also
have a transitive sense, "to lift", i.e., "to support". Here it may be an active participle in .0,
with sag being its historic patient or incorporated object: "the one who lifts the head". The
usual translation is "supporter, sustainer, champion (of)". It is translated by Sollberger as:
"protector, patron (literally, '(he who) supports the head')".
Lesson 13 149
Text 1 3 a
* D *r-@ :::[!]
.
Ir " -f
.._ -
. -.
.- .. .
<9 'J
r:li1 ,
*" r--'fi 111-
111- )ECf
<J-- <> @hj
11
-"4
Cif.- r -h1
firFl * J7J . . ..- r - -I
.
...
-
-
- -
i . . . .-
.I,' I , Ij (!!:>
I.b, /iil :> t:t:::J
- I --- ,I f'
' *
=
There is a certain amount of variation in the cuneiform signs in this text. The da at the
end of line 4 has an initial vertical, not seen in the da in line 9. Similarly, the sign read nitab
in line 7 has a vertical not seen in the sign read us in line 5.
150 Lesson 13
Text 1 3b
Lesson 13 151
construction". It is fonned by deletion of the entire verbal prefix chain, and nominalization
of the remaining verbal root in .a (except in certain cases, not discussed here).
To express "Amar-Sin, whom Enlil proposed in Nippur" using the full form of the
relative clause, would be approximately: Amarzuen [Enlil.e Nibru.a mu.mu.n.pad.0].a.
(The first Imul is the nominal component of the compound verb mu ...pad: the second
Imul is the conjugation-prefix.) To express the same idea using the reduced relative
clause, the prefix chain is deleted: Amarsin [Enli1.e Nibru.a mu.pad.0].a, which is the fonn
in Text 13. The mu which remains is the nominal element of the compound verb, not the
conjugation-prefix. Since the nominal component of a compound verb is not part of the
prefix chain, it is not deleted.
Several English translations of this construction are possible: "Amar-Sin, whom Enlil
proposed in Nippur", "Amar-Sin, proposed by Enlil in Nippur", etc.
Both full and reduced relative clauses are common in Sumerian, but it is not known if
there are rules governing their distribution. Certain fonnulaic expressions tend to prefer
one construction, while other fonnulaic constructions prefer the other. For example, "the
man who built ... " always appears in these texts as lu in-du-, but "the man proposed by
Enlil" always appears in these texts as lu dEn-lil-Ie mu-pad-da.
5. sag-us _dEn-lil-ka sagus.e.Enlil.(a)k.a(k), "patron of the temple of Enlil". The geni
=
tive phrases seen till now have consisted of two nouns or nominal phrases. However, it is
also possible to have a genitive phrase consisting of three or four elements, such as "the
king of the temple of Nanna". Such genitive phrases are fonned by the addition of an extra
".ak" for each new element in the genitive phrase. For example, "the temple of Nanna" is:
e.Nanna.k; "the king of the temple of Nanna" is: luga1.e.Nanna.k.ak; "patron of the temple
of Enlil" is: sagus.e.Enli1.ak.ak. (Sequences of two ".ak"s are sometimes referred to as
"double genitives".)
Sequences of four nouns or nominal phrases (therefore, with three ".aks"), although
pennissable, are uncommon. Sumerian does not seem to tolerate a sequence of more than
three ".ak"s; if such a situation would arise, no more than three are used. More commonly,
a periphrasis of some kind is used instead.
The genitive phrase in this line is written sag-us-_ dEn-lil-ka. The expected la! of the
first genitive marker does not appear in the writing. One might have expected a writing
such as sag-us-_dEn-lil-Ia-ka, or some such. A similar writing occurred in Text 1 1 : bi-li
nam-munus-ka-ni, for bili.nam.munus.(a)k.ani. As discussed at length in that Lesson, it is
not known whether the problem is at the orthographic, phonological, or morphological
level. Falkenstein's school, for example, would read the lil-sign here as lila2 : sag-us-
dEn-lil!!z"ka.
Discussion: structure
Text 13 is a standard inscription, similar to Text 7. It consists entirely of a series of
appositives, serving as epithets to the name Amar-Sin in line 1 ; there is no finite verb form:
Amar-Sin,
proposed by Enlil in Nippur,
Lesson 13 153
by the detenninative for place: En-lilki . That is, the writing does not attempt at all to repro
duce the phonetic sequence INibru/. Rather, the Sumerian reader would understand the
written signs as standing for "the place associated with the god Enlil", that is, Nibru.
Transliterations of such place names (and of similar common nouns) vary. Older
practice tends to reproduce the basic value of the signs fonning the word: Ep._lilki . More
current practice is to use the name of the GN (assuming it is known) : NibrukI.
-Relative clauses
The tenn "participial construction" has been used to describe constructions such as in
lines 2-4, because a reduced verbal fonn with a nominalizer is fonnally identical with what
has been called here a passive participle; a reduced mu-n.-du, becoming du-, is fonnally
identical with the passive participle du-!.. In origin, in fact, passive participles are all
probably reduced relative clauses, in special syntactic environments.
The construction called here "reduced relative clause" is often referred to as the
"Mesanepada (or Mesannepadda) construction". (This name for the construction
apparently goes back to Falkenstein.) Me-an-ne-pad-da was the founder of the First Dy
nasty of Ur ("Ur I") sometime around 2550 BC. His name means "the young man whom
An chose", or "the young man chosen by An": Me.an.e pad.a. This is the minimal fonn
of the construction: a head noun (mM); an agentive marked in .e (An-; a verbal root
(pad); a nominalizer. (A few other names of the type "X-an-ne-pad-da are also known.)
Lines 1-4 of Text 13 are only slightly more complicated; Text 13 includes a locative phrase,
and also uses a compound verb (mu ...pad) instead of just pad.
Other examples of this construction occur in Gudea. A temple is referred to as: E
ninnu An-ne ki-gar-ra, "the Eninnu temple, which An established (literally, 'placed on the
ground', ki.a)", or "the Eninnu temple established by An".
The only study that deals specifically with relative clauses in Sumerian is by Gragg
(1972a). It was written for a non-Sumerological audience, and is by design short and
schematic, but has several useful observations. He points out that the syntax of relative
clauses, particularly reduced relative clauses, is not completely understood. Henri Limet
(1975a) has studied the parallel use of .a in participial and relative sentences.
- Standard inscriptions
Copies of this text have been found at several different sites, including Adab, Bad
Tibira, Eridu, Girsu, Isin, Kisurra, Sippar, Tell el-Lahm, Ur, and Uruk. All copies were
inscribed on bricks. This situation is not uncommon with standard inscriptions, which can
be found anywhere the ruler held sway, or where building-activity was conducted under his
aegIS.
- History
Shulgi apparently died of old age; he had ruled for some 47 years. His son, Amar
Sin, ruled only nine years, and not much is known of his activities.
Lesson 13 155
Text 1 3c
supplementary
Another brick.
Lesson 1 4
This text is a brick building inscription of Amar-Sin; it is essentially an expansion of
Text 1 3.
Sign-list and vocabulary
Notes
En-ki The god of the subterranean waters, and also the god of wisdom. He was a son of
Nammu. His name apparently means "lord of the earth", en.ki.(k). There are spellings
which show that this name is a genitive phrase, not a noun-noun compound.
Why a god whose name means "lord of the earth" became associated with water is not
entirely clear; Jacobsen has spoken of "the role of water in fructifying the earth". It has
also been speculated that the element Ikil appearing in this name is a different word than
the word lkil meaning "earth", perhaps the same ki appearing in the compound verb
kig2
The Sumerian god Enki was equated with the Akkadian god Ea <E.-. The name of
the latter is of uncertain etymology; it does not inflect for case. In the bilingual lists from
Ebla, the Ebaite equivalent of Enki is written E-!!9 ' This would appear to be an inflected
form of the name, with the nominative case-marker. It has also been speculated that the
Akkadian writing E- and the Eblaite writing E-!!9 are actually phonetic spellings repre
senting a Semitic form something like Ibayyu/, "the living one". This idea is explicitly
developed by Cyrus Gordon (1987:19-20).
abzu This is composed of two signs, the zu-sign followed by the ab-sign. However, it is
known that the ab-sign was read before the zu-sign. That is, the word was pronounced
something like labzu/. The phenomenon is similar to that of the DN Zuen, which was
pronounced lzuen/, although written en-zu. In older transliterations it may appear as zu
ab.
The original use of this term was mythological. It referred to the subterranean fresh
waters, which the Sumerians believed lay below the surface of the earth. These waters fed
the wells, streams, rivers, marshes, etc. These waters were the special purview of Enki.
The term abzu was also used as the name of a large temple in Eridu, built to honor
Enki. Most of the work on this temple was done by Amar-Sin, although it was his father
who actually began the construction. This temple apparently stood over a fresh-water
lagoon.
The term was later applied to a cultic object, presumably some kind of water-basin
157
158 Lesson 14
used in the temple. At a number of sites in Mesopotamia, objects have been found which
archaeologists have identified with the term abzu. Several such objects have recently been
found at Ebla.
This word was borrowed into Akkadian as apsfi, glossed by the CAD as: " 1 ) deep
water, sea, cosmic subterranean water, 2) (a personified mythological figure), 3) water
basin in the temple".
It is possible that the Eblaite equivalent of Sumerian abzu and Akkadian apsfi appears
as -ba-si (Fales 1984: 1 84), but the interpretation is somewhat uncertain.
The English word "abyss" is thought to derive from this Sumerian word, via
Akkadian and Greek. It has been speculated that the word is not native Sumerian, but
rather derives from a substrate language.
Lesson 14 159
Text 1 4
-t Jtt Tt
)-> z
S:j
ffiPT
'"* :f-:i
:d gg l
I
r!j\
160 Lesson 14
Irl of the dative case-marker was completely lost, producing /lugalnil for the dative.
However, adoption of such a view entails rather baroque convolutions in explaining such
forms as h!gal--ni-ir. Falkenstein, in fact, speaks of "eine sekund11re Restitution" of the
dative case-marker Irl, reflecting a period when Sumerian was beginning to fall out of use
as a spoken language. This would mean that the original form was Irl, but then the Irl
was lost, and later it was "restored". However, such an explanation encounters strong
linguistic objections, and also historical objections. Also, it does not account for the rise of
explicitness seen in other areas of the grammar. In Text 16, for example, there occurs mu
na-an-du, the first occurrence in the body of texts studied here where the personal-affix
cross-referencing the bamt!!-agent is actually written.
1 2. Because of the ambiguity of the term abzu, it is not clear what the patient in this line
refers to: the temple built by Amar-Sin and his father, or to a cultic object within this
temple.
Discussion: structure
The structure of this text is:
The first nine lines are essentially the same standard inscription seen in Text 1 3. The
rest is a straight-forward building inscription. The result is that this inscription begins with
the name of the king (the agent) instead of with the name of the deity (the benefactive). It is
as if the scribe began with a stock standard inscription, then tacked on a building
inscription.
There are two differences between Text 1 3 and Text 14. The first is in the order of
lines 2 and 3. In Text 1 3, the locative phrase within the reduced relative clause precedes the
agentive phrase, but in Text 14, the agentive phrase precedes the locative phrase. Text 14
uses the more usual syntax. In Text 1 3, there is presumably some emphasis on the word
"Nippur". The second difference is in line 7. Text 1 3 uses a title which occurred in several
other texts: nitab kalag-g. Text 14 uses lugill kalag-g,a.
- Orthography
On the one hand, writings which explicitly represent the dative case-marker Irl may be
viewed as part of the general process of the Sumerian writing system becoming more and
more explicit in its representation of phonological and morphological features. On the other
hand, it is hard to understand exactly how such a practice appeared - what motivated a
scribe, practicing by its nature a conservative craft, to write the Ir!?
It is not known when Sumerian began to die out as a spoken language; many scholars
162 Lesson 14
believe that such a process was already on-going during the Ur III period. If so, the
increase in explicitness in the texts - such as the writing of the dative case-marker Irl -
may be correlated with an increased need of the scribes for help in reading and writing
Sumerian. That is, as the scribes' knowledge grew more and more "shaky", there was a
need to write the morphemes down in an unambiguous way. At the same time, there may
have been felt a scholastic tendency to write all morphemes down. Similarly, in our own
scholastic tradition of transcribing S umerian, we are prone to write down full underlying
forms of morphemes.
-Ergativity
There are two ergative case-markers in this sentence: the .e in line 9, marking the
agent of the main verb in line 1 3, and the .e in line 2, marking the agent of the verb in line 4
(embedded in a relative clause). Potentially, this could cause a certain amount of confusion.
If one thought that the .e in line 2 marked the agent of the verb coming up in line 1 3, the
text would start to become rather confused. In practice, however, the formulaic nature of
these texts helps to prevent such confusion. In the spoken language, there were probably
features such as stress and intonation which helped obviate such problems.
- Textual problems
At the end of line 9, the autograph reads ke4 ; the ke4 -sign combines the Ikl of the
genitive marker with the lel of the ergative case-marker. This particular use of the ke4 -sign
has occurred in several of the previous texts.
However, at least two of the three other published exemplars of this text read a ka
sign, not a ke4-sign, at the end of line 9. This is hard to explain. There is no evidence for a
(phonological) change of le/ ) lal at this period; sporadic cases do occur in later S umerian,
but under different conditions.
One possibility is to see a long anticipatory genitive, of a kind not seen previously.
The essence of the sentence would be:
"Of Amar-Sin ... , to Enki his beloved lord" )
"To Enki, the beloved lord of Amar-Sin".
Although somewhat similar anticipatory genitives do occur in Sumerian, there seem to be
no exact parallels to this construction. Also, such an interpretation would not leave any
overt agent for the finite verb.
It is difficult to find a satisfactory explanation for the writing with the ka-sign. It is not
simply a scribal error, since the ke4 -sign seems to occur in only one exemplar; two
exemplars clearly have the ka-sign, and one exemplar is slightly damaged at the crucial
point.
The fact that the ke4 -sign is apparent in only one copy is in itself suspicious. It is
possible that this writing represents an (unconscious) attempt by the scribe to bring the text
into line with more common Sumerian morphology and orthography. (This assumes, in
fact, that the ke4-sign is indeed present, and that it is not an error on the part of the modern
day editor of the text; there is no photograph available of the text.)
Lesson 14 163
-Titulature
-Terminology
Text 1 4a
supplementary
alam statue
mu name
ama mother
sig9 (g)
1] to be narrow
1'- me
4t- Im
eb
Notes
alam Transliterated by some as alam, and by others as alan. Some Sumerologists believe
that alan is the older, and alam the later form. On the other hand, Lieberman reconstructs the
original form as /alag/; he believes that nasals in word-final position were neutralized with
respect to point of articulation. Others have postulated some kind of connection with the
word appearing in Akkadian as almu, in Hebrew as elem, "image".
barag-sig9 -ga The exact meaning is unsure. It may mean "narrow dais", barag.sig.a. It
is translated by the PSD as "socle (of a statue)". It was borrowed into Akkadian as
barasigQ, translated by the CAD as: "low socle for cultic purposes".
1 65
1 66 Lesson 15
kur This was equated with Akkadian aml and nakaru. Both have many meanings. For
nakaru in the D-stem, the CAD lists, among others, " ... 8. to discard an object (tablet, stela,
etc.), to remove an inscription ... 9. to clear away rubble, etc., to discard, remove from a
container, to demolish a building ... 1 1 . ... to place an object in a new location ... ".
Most frequently, kur governs a direct object (patient).
bU6 This is translated by the PSD as "to tear out", "to pull out", "to uproot", "to extirpate".
Although the meaning is clear, the precise reading is not. The sign is a ka-sign with an
inscribed kar-sign. This is most clearly seen in the Neo-Assyrian form of the sign: .
Sumerologists often transliterate such inscribed signs with an "x": KAxKAR.
Older works transliterate the ka-sign as bus (when necessary), KAxKAR as and
KAxSU as bu. Recent works, however, transliterate both KAxKAR and KAxSU as bu,
and add the fuller transliteration in parentheses: bu(KAxKAR); bu(KAxSU). However,
such a system can lead to confusion, because it is easy for the forms in parentheses to be
accidentally dropped in the mechanical process of printing. And, since the entire function
of transliteration is to provide a one-to-one correspondence of a specific cuneiform sign
with a specific transliteration, it seems counter-purposeful to use bu for two different signs.
Therefore, the older procedure is followed here, and the sign is transliterated as .
It is probable that the root of this word ended in some kind of Ir/-Auslaut; in Text 1 5
i t appears i n a verb form written i-bu 6 -re-. However, no Iburl value for this sign is re
cognized by the standard sign lists. A possible reading buzurs is recorded, but this may be
some other use. The problem deserves further study.
bU6 usually governs a direct object (patient).
til This sign has several meanings in Sumerian. In its reading as til, it is equated with
Akkadian gamaru, laqatu, and qatu. The CAD glosses qatu as " 1 . to come to an end, to be
used up, 2. to perish, 3. to become completed, finished, settled". In the causative stern,
uqtu is glossed as "to bring to an end".
Lesson 15 167
Text 1 5
* L:) f-*- EE
rf
M .ti
<H) A
$=I , :J
* mE1
lsWT> Fr1 1 The sign is partly restored from the
d plicate, No. 90039.
4
168 Lesson 15
*
-- .
x if\:f rt "
] rwti> " av=r
ii===> 1N n p
. ).
* } -<fPJ
I ,
T=:I m
* f-
Lesson 15 1 69
Notes
11 9: The nam-sign and the U1!-sign are either partly effaced or poorly drawn on the
original. Not all scholars, in fact, believe that there even is a U1!-sign in this particular text.
But since this curse-formula occurs in several other texts, its restoration is relatively certain.
Transliteration Transcription
I 1: dAmar-dZuen Amarzuen
2: NibrukC1! Nibru.a
3: dEn-lil-Ie Enlil.e
4: mu-pad-da mu.pad.a
5: sag-us sagus
6: _ dEn-lil-ka e.Enlil.(a)k.a(k)
7: lugill-kalag-g1! lugal.kalaga
8: ' ki-ma
I !wI -Vnm
...! lugal.Vrim.a(k)
S
9: lugal-an-ub-da-lirnrnuz-ba-me lugal.anub.da.(k) limmu.bLa(k).me.(en)
10: alam-ba alam. bLa(k)
1 1 : dAmar-dZuen ---.! ki-ma Amarzuen kLaga.a. Vrim.a(k)
k' -1!g-!!TVnms
12: mu-bi-im mu.bLm
13: alam-ba alam. bLa(k)
II 1: lu ki-gy..b-ba-bi lu kLgub.a.bL0
2: i b-da-ab-kur-re-!! Lb.da.b.kur.e.0.a
3: barag-sig9-g1!-bi baragsiga. bL0
4: i-bU 6-re-!! L(b.)bur.e.0.a.(d)
5: dNanna Nanna
6: lugill-Vrimskt ma-ke4 lugal.Vrim.ak.e
7: dNin-gill Ningal
8: Ski-ma-
ama-Vnm k 4 ama. Vrim.ak.e
9: nam-U1!-ba-an-da-kur, ne namue.ba.n.da.kur.(e.e)ne.0
1 0: numun-na-ni numun.ani.0
1 1 : h-eb-til-Ie-ne ue.(L)b.til.e.ene.0
Translation
I 1: I am Amar-Sin,
2-4: proposed in Nippur by Enlil,
170 Lesson 15
The distribution of -me - -me-en does not exactly correspond to a difference in time.
In older Sumerian, -me-en is the norm, but in Ur Ill, both -me and -me-en occur, with -me
predominating. However, both also occur in later texts. It is not sure if this should be
regarded as an orthographic or as a phonological problem; it is discussed further below.
The regular form of the enclitic copula for the third person singular is laml, written
with the am-sign. After a vowel, the enclitic copula appears as Iml, as in line 1 0 of Text
1 5.
1 0. The next three lines give the actual name of the statue. The construction is an
anticipatory genitive, with an enclitic copula: "of this statue, its name is ... ". Thus, the bi of
mu-bi resumes alam: alam.bi.a(k) ... mu.bi.m.
1 1 . The actual name of the statue is: "Amar-Sin is the beloved of Ur". This is an
equational sentence, and so one might have expected to find an enclitic copula. However, it
is not uncommon to find simple equational sentences without a copula, and this name may
be such an instance. It is also possible that the name is not a complete sentence, but rather
is a noun phrase with an appositive: "Amar-Sin, the beloved of Ur"; this is discussed
Lesson 15 171
further below.
The next few lines are rather complicated. Lines I: 1 3 through IT: 4 are all the
comitative complement of the verb in IT: 9. This complement includes a relative marker iliD
governing two relative clauses. The first is marked by the - at the end of line IT: 2. The
second is marked by the -ft at the end of line IT: 4. All of this is embedded inside an
anticipatory genitive:
"of this statue, the man who changes its position and tears down its pedestal" )
"the man who changes the position of this statue and tears down its pedestal".
The anticipatory genitive in I: 13 is resumed by the -bi in IT: 1 and 11: 3.
11 1 . ki-gub-ba = ki.gub.a, "standing place" or something similar. Presumably, gub-ba
-
i . b . da b . kur . e . 0 a
(1) (2) (3) (4) (5) (6) (7) (8)
(1) conjugation-prefix
(2) optional pronominal-prefix
(3) comitative dimensional-prefix
(4) personal-affix cross-referencing patient
(5) verbal root
(6) maru-suffix
(7) personal-affix cross-referencing agent
(8) nominalizer
The entire clause, nominalized in -!!, stands in apposition to ill, forming a relative
clause.
4. i-bu6-re-!! = L(b).bur.e.a. i is the conjugation-prefix. However, the verb has no dimen
sional-prefix, nor (on the surface) does it have any personal-affix cross-referencing the
apparent direct object, baragsiga.bi.0. Since the verb form is in the maru, one might have
expected to find /b/ immediately preceding the verbal root. There are at least two possible
reasons for its apparent absence. Lines 3-4 are roughly parallel in form and content to lines
1 -2. It is possible that its use in the verb form in line 4 would have been felt as redundant;
that is, the parallelism in construction permitted deletion of the personal-affix .b. Some
scholars have suggested that personal-affixes are only present (and therefore only written)
in ambiguous contexts.
However, there may be an entirely different reason for its absence. Perhaps there was
a phonetic reduction of /ibburea/ ) /iburea/. That is, the problem may be phonological, not
morphological. As is often the case in Sumerian, it can be difficult to determine whether a
problem is orthographic, phonological, or morphological in nature.
4. The assumed comitative case-marker .(d) at the end of this line marks lines I: 13
through IT: 4 a s the comitative complement o f the verb nam. . .kurS i n line 11: 9. The same
use of the assumed comitative with this verb occurred in Text 1 2.
6-8. The two agents of the verb forms in lines 9 and 1 1 are both marked by the ergative
case-marker .e.
9 . .nda is the comitative dimensional-prefix plus the optional pronominal-prefix .n. Here
the animate form is used, because it refers back (essentially) to "the man who ... ".
All the verb forms seen up to this point have been singular. Here, there is a plural
agent ("Nanna and Ningal"), and so the verb must be put into the plural (Sumerian has no
dual). The plural third person of a maru verb is usually written with a suffixed --ne.
S umerologists have different understandings about the morphology implied by this writing.
One analysis sees this writing as reflecting: e.ene.0. The .e is the mat-u-suffix; .ene is the
plural marker; and .0 is the personal-affix cross-referencing the agent. Another analysis is
Lesson 15 173
to isolate the morphemes of this ending as: e.0.ene. A third analysis is to read the ne-sign
as de, reflecting quite a different understanding of the morphology. This is a very difficult
issue to resolve. Here, the first analysis has been followed.
A problem in this particular verbal form is the fact that only -ne is written, not --ne.
How should this writing be understood?
The Falkenstein school would read the first sign with an overhanging vowel, that is,
kure2' A variant of this solution is to read the kur-sign as kUfUS : kUfU s-ne would represent
kur.e.ene.0. The second lul in the sign kUfUS would represent the assimilation of the un
derlying lel to the first lu/. There are numerous other cases where the marii-suffix .e is
assimilated to an I ul-vowel of a verbal root.
But has been stated several times, the mnemonic nature of the script may have meant
that there was no need for the scribes to write down the full forms of the morphemes.
To summarize the verb form:
of animates and inanimates), and here numun was probably felt as a collective.
Since til is a transitive verb, b is used with the maru form of the root. til is a member
of the affixation class, so its maru-root is ti1.e. The verb is written as expected, til-Ie-ne.
To summarize the verb form:
be i b til e ene o
(1) (2) (3) (4) (5) (6) (7)
(1) modal-prefix
(2) conjugation-prefix
(3) personal-affix cross-referencing patient
(4) verbal root
(5) maru-suffix
(6) plural marker
(7) personal-affix cross-referencing agent
Discussion: structure
This text is composed of three sentences:
I: 13-II: 4 accompaniment
11: 5-6 agent 1
11: 7-8 agent2
11: 9 verb
11: 1 0 patient
11: 1 1 verb
- Sign formation
In Lesson 3, it was pointed out that some cuneiform signs are in origin combinations
of a pictographic sign with an inscribed phonetic indicator of some kind. The ama-sign is
thought to be one of these. It has an inscribed digir-sign, one of whose phonetic readings
is am6 . Therefore, it has been suggested that this component of the sign is a clue to the
Lesson 15 175
pronunciation. Unfortunately, it is difficult to make this square with the fact that the rest of
the sign appears to be the pisan-sign, which basically means "box" of some kind. It has
also been proposed that there is some obscure symbolism involved, "mother" being
represented as a "divinity" within a "box"!
- Loan words
In Text 6, the word barag occurred; this term was borrowed into Akkadian as parakku.
The amissable /g/-Auslaut shows up in Akkadian as a geminated voiceless consonant.
barag-sig9-gi!., however, was borrowed into Akkadian as barasigu. The amissable / g/
Auslaut of the barag element, here in syllable-final position, does not appear. However, the
intervocalic / g/ of gg-gi!. (sig.a) remains. The word-initial /b/ of barag is treated dif
ferently in each loan word. In parakku it is reflected as /p/, but in barasigfi it is reflected as
/b/. This shows that barasigu is a later borrowing into Akkadian than parakku. It is us
ually assumed that in relatively older loan words from S umerian, Sumerian voiced stops are
reflected in Akkadian as voiceless stops. In relatively later loan words, the same Sumerian
voiced stops are reflected as voiced stops. Needless to say, not enough is known about the
phonetics of either Sumerian or Akkadian to explain exactly what has happened. However,
such differences are one way that the entry of loan words into another language can be
dated relative to each other.
- Conjugation -prefixes
In the model of the Sumerian prefix chain used throughout this book, the use of the
conjugation-prefix is obligatory; a conjugation-prefix is present in every finite verbal form.
Therefore, in line 11: 1 of Text 1 5, it is assumed that a conjugation-prefix 1 has assimilated
into the modal-prefix b. Similarly, in Text 22, this view assumes the presence of a
conjugation-prefix 1 after the cohortative modal-prefix g.!!, although the verb form in Text
22 is written g.!!-an-!!-il.
Although both the desiderative modal-prefix b and the cohortative modal-prefix g.!!
are frequently followed by such conjugation-prefixes as mu and ba, there appear to be no
instances of writings of the type *b-l or *g.!!-l. If the 1 is in fact present, one might expect
to find at least a few occurrences of it being written (to judge by similar phenomena in the
script). However, none apparently occur. This means that the assumption that the attested
spellings all represent assimilation may not, in fact, be correct.
Jacobsen, for example, believes that the conjugation-prefix 1 is "incompatible" with the
cohortative modal-prefix g.!!. That is, the semantic information conveyed by 1 does not
permit it to co-occur with g.!!. This may mean that at times the modal-prefix g.!! is followed
by no conjugation-prefix. The up-shot is that not every finite verbal form contains a
conjugation-prefix.
This problem cannot be resolved here. However, it should be kept in mind that the
general principle stated in this book - that conjugation-prefixes are obligatory - may need
modification, particularly in the case of certain modal-prefixes.
176 Lesson 15
- Pronominal-prefixes
The form of the first person singular was probably not simply .0. Some irregular
writings have led scholars to speculate that the first person singular marker had either a
vocalic component (perhaps lel, similar to that of the second person), or a consonantal
component (perhaps a semi-vowel or glottal consonant). The fonns of the plural are much
less clear.
If more than one dimensional-prefix occurs in a verbal prefix-chain, it appears that
only the first dimensional-prefix can have an optional pronominal-prefix.
An older view of these pronominal-prefixes was that they were obligatory, and that
their relative infrequency in texts is due to vagaries of orthography. In this view, for
example, every .da dimensional-prefix which cross-references a third person should be un
derstood as: (n.)da. Thus, the verb fonn in Text 1 2, nam-b-ba-da-kur5 -, should be un
derstood as: nam.be.ba.(n.)da.kur.e.0, instead of: nam.he.ba.da.kur.e.0. However, this
view is not as widely accepted today, because the actual number of cases where an .n or .b
appears before an appropriate dimensional-prefix is much less than the number of cases
where they do not appear.
-Dimensional-prefixes
In line 2, the verb kur was used with the dimensional-prefix -da, although the sentence
contains no nominal phrase in the comitative case. In some cases, verbs have become
lexicalized with certain dimensional-prefixes; that is, the verb will frequently (sometimes
always) have a certain dimensional-prefix, even if no corresponding nominal phrase occurs.
Occasionally a noun phrase will be marked with a certain case-ending, but the
dimensional-prefix used in the corresponding verbal phrase will be different than e,xpected.
For example, a nominal phrase in -e (the tenninative) may be resumed by the dimensional
prefix nonnally used for the locative (-ni-). In some cases, this may have happened
because of a historical change in the rection of a noun phrase. That is, at one time a noun
phrase may have been marked by one particular case, but in time the case which was used
changed. However, the dimensional-prefix, being more closely bound, did not change.
This can lead to such cases as a tenninative being cross-referenced by a locative. In the
case of kur, for example, its complement may originally have been in the comitative case (or
perhaps the ablative case), but it eventually shifted to the absolute case. However, it carried
along its dimensional-prefix, resulting in such cases as line 11: 2, where the verbal prefix
chain has both a dimensional-prefix and a personal-affix, yet there is only one nominal
Lesson 15 177
-Enclitic copula
The fact that the first and second persons of the enclitic copula occur both as -me and -
me-en admits of several possible interpretations. The problem may have been orthographic:
word and syllable-final nasals often seem not to be written in S umerian (cf. mu-na-du for
mu.na.(n.)du). It may reflect the phonetic process of Sumerian dropping word-final nasals;
in this case, the forms written with -en are to be regarded simply as morphographemic or
archaic writings. It has also been posited that the writings -me and -me-en represent one
pronunciation, /me/. The original word-final nasal was lost, producing a nasalized vowel,
which the script could not well represent.
- Proper names
Votive objects of all kinds were often given names. This has been studied by Gelb,
who points out that such names are often complicated sentences in their own right. For
example, a statue of Gudea dedicated to Ningirsu was given the name (Gelb's translation):
"Ningirsu, -the-king(=god)-whose-heavy-might-the-world-cannot-bear, -has-decided-good
destiny-for-Gudea, -who-built-this-temple" (1956:66).
- History
In addition to the copy of the text reproduced above, a late, Neo-Babylonian copy from
the seventh century BC is also preserved. It was inscribed on what was apparently a model
pedestal. This copy is interesting because of the presence of several errors in the Sumerian.
Also, it has a colophon written in Akkadian, which seems to say that the model was to be
used in an "exhibition" (tamartu) of some kind. The page following is a photograph of two
sides of this Neo-Babylonian version of the inscription.
The following quotation from Woolley describes this object, and also says something
178 Lesson 15
$ hH honey
JP i oil, fat
rv i-nun butter
.
vme, wme
.
siskur2 sacrifice
Wig to cease
!!. and
nu not
Notes
11\1 It is thought that the bee is native to Syria, but not to Mesopotamia. So Iftl (and its
Akkadian equivalent dgpu) may be a Syrian import. It has also been speculated that hU and
dispu were actually a kind of syrup made from fruits.
i-nun Literally, "oil of the prince". Some scholars interpret this as "butter", others as
"ghee" (that is, clarified butter). The latest discussion (K. Butz, 1973-74:37) interprets it as
"Butterschmalz" ("butterfat"). The Akkadian equivalent of i-nun is bimetu; this is trans
lated by the CAD as "ghee".
On the surface, i-nun would appear to be a noun-noun compound, of the type dis
cussed in Lesson 9. However, there are spellings in other texts which show that this is a
genitive phrase, "oil of the prince", Lnun.a(k). For example, in Gudea this expression is
written both i-nun and i-nun-na. The term i-dug-nun-na, = i.dug.nun.a(k), "good oil of the
prince", also occurs, but it is not sure what this means.
In older works, the word for "oil, fat" is transliterated as hi, instead of i.
gdtin The word for "wine" also appears as tin. gestin, therefore, may originally have
181
1 82 Lesson 16
meant "wine-tree, vine, grape", etc.; the meaning then became extended to "wine". The
Akkadian equivalent, karanu, is glossed by the CAD as: " 1 . wine, 2. grapewine, 3.
grapes".
siskur2 This word is normally written just as it appears here, by graphic reduplication of
one sign. This sign appears to be the amar-sign inside of which is the barley-sign, Se.
Thus, the original, pictorial significance of the sign may have been "grain-fed cattle", or
something similar.
Sometimes the word Isiskurl is written with only one, instead of two, signs. In this
case it is transliterated as siskur. Earlier Sumerologists thought that the writing with one
sign was a singular, and the writing with two signs was a plural; this means that the writing
with two signs was sometimes transliterated as siskur-siskur. The word is further dis
cussed in Lesson 19.
The pronunciation of the sibilants is unsure, and so it is also transliterated as siskur2 '
sizkur2' Siskur2 , etc.
ilig Or, silig. This root is very uncommon in verbal forms; it is mostly used in participial
or infinitival constructions.
Lesson 16
1 83
Tex t 1 6
184 Lesson 16
Transliteration Transcription
1: dEn-lil Enlil
2: lugill-kur-kur-ra luga1.kur.kur.a(k)
3: lugill-ki-.M-M-ni-ir luga1.kL aga.a.anLr
4: dAmar-dZuen Amarzuen
5: dEn-lil-Ie Enli1.e
6: NibrukC Nibru.a
7: mu-pad-da mu.pad.a
8: sag-us sagus
9: _dEn-lil-ka e.Enli1.(a)k.a(k)
1 0: lugal-kalag-g luga1.kalaga
I !g-
1 1 : ...! ' ki-ma
I Unms luga1.Urim.a(k)
12: lugill -an-ub-da-limmurba-ke4 luga1.anub.da.(k) limmu.bLak.e
13: lal i-nun e lal inun
14: y getin u getin
15: ki-siskurTra-ka-na kLsiskur.ak.anLa
16: nu-i1ig-g nu.i1ig.e(d).0.0
17: mu-na-an-du mu.na.n.du.0
Translation
1: For Enlil,
2: king of the lands,
3: his beloved king -
4: Amar-Sin,
5: proposed by Enlil in Nippur,
8: patron of the temple of Enlil,
10: the mighty king,
1 1 : the king of Ur,
1 2: the king of the four quarters -
17: built -
13: the temple (where) honey, butter
14: and wine
15: in his place of sacrifice
Lesson 16 185
mate personal-affix .n, cross-referencing the agent in the bamn!, appears explicitly in the
script.
There is no dimensional-prefix cross-referencing the locative nominal phrase in line
15.
Discussion: structure
The structure of this text is:
[Enlil, lugal.kur.kur.a(k), lugal.kLaga.a.ani].r benefactive
[Amarzuen, Enlil.e Nibru.a mu.pad.a, agent
sagus.e.Enlil.(a)k.a(k),lugal.kalaga,
lugal. Urim.a(k),
I ugal.anu b.da. (k)limmu.bLak].e
[e lal inun u getin kLsiskur.ak.anLa patient
nu.ilig.e(d).0].0
mu.na.n.du.0 verb
Relative clauses consisting of a nominal sentence, such as lines 1 3-16, are not
common in Sumerian, and so there is some question about the analysis. It is possible that
these lines should be understood as the name of the temple: "Amar-Sin built a temple
(whose name is) 'honey, butter and wine in his place of sacrifice shall not cease"' . In such
constructions, Sumerian does not always use the word for "name" (mu); instead, it conjoins
the name directly, as an appositive. An example occurs in Text 19: bad-Mar-tu Mu-ri-iq
Ti-id-ni-im, "the Martu-wall (whose name is) 'Muriq-Tidnim" ' .
Lesson 16 187
- .ed
There is no standard term used to refer to .ed. It can appear both on finite verbal
forms, and on nominal forms derived from verbal roots. On finite verbal forms, it occurs
with both marii-roots and bamt!!-roots (although there has been much argument about this).
It seems to occur with both transitive and intransitive-passive forms. On nominal forms, it
occurs on the "infinitive", the active participle, and the passive participle.
Because the Idl is amissable, and because the lel can assimilate into other vowels, it
is not always easy to determine even if it is present or not. The assumed Id! does not show
up in Text 1 6, and it is therefore also possible that some other interpretation of line 1 6 is to
be preferred. In spite of several recent attempts to unlock its morphology and syntax (the
most recent is that of Gerd Steiner [1980]), it remains elusive.
-Participles
The traditional view of Sumerian is that it has two participles, an active and a passive
intransitive (Poebel used the terms nomen il,gentis and nomen actionis). The active ends in
.0, and the passive-intransitive in .a. Each participle can also appear with the element .ed.
This yields four possibilities:
X.0 X.a
X.ed.0 X.ed.a
However, a fair number of exceptions seem to occur. There are cases of participles in
.0 which seem to have a passive meaning, and participles in .a which seem to have an
active meaning. For example, the form in Text 1 6 is mig-g, presumably for ilig.ed.0.
Here it is being used intransitively, and therefore one might have expected a participle in .a.
The situation was undoubtedly more complex than the above table would indicate. Most
discussion of non-verbal forms in Sumerian has applied Indo-European or Semitic
grammatical categories and terms to the S umerian forms.
The most recent, full-scale, study of the participles is that of I.T. Kaneva (1970). She
analyzes the forms differently:
X.0 "transitive participle of the imperfect aspect"
X.a "transitive participle of the perfective aspect"
X.a "intransitive participle"
However, she does not apparently recognize the existence of reduced relative clauses,
whose existence complicates the picture.
- Personal-affixes
This text marks the first time that the personal-affix .n cross-referencing the bamt!!
agent appears in the writing in the body of texts studied here. As discussed in Lesson 14,
this is part of the on-going process of the Sumerian writing system becoming more and
more explicit.
188 Lesson 16
- Conjunctions
The native Sumerian conjunction meaning "and", linking nouns and nominal phrases,
is -bi-da, suffixed to the second noun: an-ki-bi-da, "heaven and earth". It is thought that bi
here is the possessive-suffix, and da is the comitative case-marker. Thus, this originally
meant "the heavens along with its earth".
Sumerian also has a conjunction meaning "and", linking verbs: -in-g, occurring on
the last verb of a series of two or more, in the position after the modal-prefixes but before
the conjugation-prefixes. No examples seem to occur in the Ur III royal inscriptions. A
late example is: ... za- in-g--zu, "you also know", for: inga.i.e.zu. Because in-g is
used so infrequently, it may have had some emphatic value, instead of straight co
ordination.
- History
This door socket (along with another, having a different inscription) was found in
what has been described as the "small shrine" in Nippur, part of a large temple complex
dedicated to Enlil. The door socket bears traces of an inscription of Lugal-kiginne-dudu,
who ruled in Uruk approximately 2400 BC (that is, some three hundred years before
Amar-Sin); this early inscription was also dedicated to Enlil. Thus, Amar-Sin's builders
did not fashion a new door socket - rather, they engraved their inscription upon this
already-extant door socket, which presumably they had uncovered in their work of
rebuilding the Enlil temple.
-Literary parallels
The combination "honey, butter and wine" occurs in other Surnerian texts. Sometimes
the three elements are listed in the same order as in this inscription; sometimes the order
varies. For example, in a literary text entitled "Nanna-Sin's Journey to Nippur", there
occurs: "May Nanna-Sin make butter, honey, and wine (i-nun lal geStin) abundant"; this
particular line is repeated verbatim five times.
The expression i-nun has recently been found in one of the Sumerian-Eblaite voca
bulary lists. Unfortunately, only the Sumerian for this particular expression is given, not its
Eblaite equivalent. One of the other entries in this vocabulary list, however, is for "good
oil". The Sumerian column is i-dug; the Eblaite equivalent is sa-rna-nu t-bu.
Lesson 16 189
Text 1 6a
supplementary
This inscription was engraved on a small bead of agate, used as a votive object.
Because of its small size and somewhat irregular surface, the signs are less than elegant.
No photograph is available.
Notes
Commentary
\9\
1 92 Lesson 1 7
o utu sun
kalam land
en priest, priestess
JP li
tt>T ta
un
o:J ab
Notes
Kar-zid-da The name means "the effective quay", kar.zid.a. Karzida was apparently a
quay at Gaesh (written Ga-eSki and Ga-eSs ki). The latter was close to Ur, and evidently
had some kind of cultic connection with it; it had its own Nanna-temple. Neither the site of
Karzida nor of Gaesh has been identified.
kalam This word is apparently only used to refer to Sumer itself. It only occurs in the
singular, and is especially common with a possessive-suffix. The original pictographic
value of the sign is unknown. This word is further discussed in Lesson 19.
gi6 -pAr This was a section of the temple complex; it was where the en-priest or en-
Lesson 1 7 193
priestess lived. It was borrowed into Akkadian as giparu. It is discussed further below.
en Besides its use in political contexts (where the conventional translation is "lord"), en
can also be used to refer to a specific kind of priest or priestess.
zid It is difficult to pin down a precise meaning for zid; it can mean something like
"someone (or something) that does what he/it should be doing". The Akkadian equivalent
is kInu, glossed by the CAD as: " 1 . true, reliable, just, 2. honest, decent, loyal, 3. correct,
normal, regular, sound, legitimate".
ul This sometimes appears as ul, as in, for instance, uru-ul, "the primeval city". However,
sometimes an lit follows, resulting in graphic reduplication: ul-H . And sometimes ul or
ul-li is followed by la/, presumably the same nominalizer seen in such adjectives as kalag
g. In Text 17, it is written: uHi-.
It is probably related to ul meaning "bud" of a flower.
kur9 In older works, usually transliterated as tu; some modern-day Sumerologists still
prefer this reading.
sud4 The verbal root meaning "to be long" can be written in two different ways in
Sumerian. The most common writing is: )---- , read as sud (or su). However, it can also
be written (as in Text 17) as: >t>- . Older works tended to understand this last sign as
gid, but most scholars now read it as sud4 (or su 1 3 )'
It can be seen that the sud4 -sign is basically just a non-gumlfied version of the sud
sign. In general, it is always possible to use a non-gunufied sign in place of a gunufied
sign. For example, the gi-sign is frequently found in place of an expected g4. In Lesson
1 , it was mentioned that the name of the city of Ur is usually spelled Se-ab, but also Se
unug. It is possible that the unug-sign is a gunufied ab-sign.
This is why >{>-- is now read sud4 instead of gid. At one time, it was thought that
Igidl and Isud! were two different roots for "to be long", but now it is assumed that they
are just variant writings of the same root I sud!.
The final consonant of the root was probably the Idr/-phoneme discussed in Text 12.
In Text 17 it is followed by the re6 -sign (with its initial Ir/).
Lesson 1 7
1 94
Te xt 1 7
Lesson 1 7 1 95
Notes
1 1, 18. In this text, the kalam-sign and the un-sign are very similar in shape. The histo
rical relationship between the two signs is complex. It is usually assumed that in origin
they were formally distinct. In carefully executed texts, they are still distinguished through
the Old Babylonian period. For example, in the Code of Hammurapi, the kalam-sign ap
pears as: , and the un-sign appears as: Iffif- -t .
However, the un-sign (more properly, yg; its value was either lugl or lugu/) means
"people", and there is a close semantic connection between "people" and "land". This
might mean that the signs were not formally distinct in origin, but perhaps became
secondarily differentiated.
By the Neo-Assyrian period, the two signs fall together into one sign. There are
several Neo-Assyrian signs which represent the conflation of two or more signs which in
origin were different signs. For example, the Neo-Assyrian ku-sign is a continuation of
several different box-shaped signs, which have all fallen together in shape.
17. The nu-sign is poorly drawn; a better drawn version appears in line 1 8. There appears
to be an extra horizontal stroke, running into the top of the du-sign.
Transliteration Transcription
1: dNanna Nanna
2: Kar-zid-da Karzida.(k)
3: lugal-ki-g-g-ni-!r lugal.kLaga.a.ani.r
4: dAmar-dZuen Amarzuen
5: dEn-lil-Ie Enlil.e
6: NibrukCi! Nibru.a
7: mu-pad-da mu.pad.a
8: -us sagus
9: _dEn-lil-ka e.Enlil.(a)k.a(k)
10: illgiI-zid digir.zid
1 1 : dUtu-kalam-ma-na Utu.kalam.anLa(k)
12: lugal-kalag-gi! lugal.kalaga
13: ...!
I !gfL- ' 5ki-ma
I Unm lugal.Urim.a(k)
14: lugal-an-ub-da-limmuzba-ke4 lugal.anub.da.(k) limmu.bLak.e
15: Kar-zid-da-i! Karzida.a
16: ud-ul-li-i!-ta ud.uli.a.ta
17: -par-bi nu-du-am gipar.bi.0 nu.(i.)du.0.a.am
18: en nu-un-til-Ia-am en.0 nu.(i .)n.ti1.0.a.am
1 96 Lesson 1 7
17. All the verb forms seen up to this point have been in what we would call in English
the "active voice". There has been a long discussion about whether or not Sumerian has a
"passive voice". Some Sumerologists say that Sumerian has no passive voice. Others say
that Sumerian is basically passival in nature.
To some degree this is a question of linguistic theory, and not of Sumerian. It is a
question of the definition and nature of active and passive, of the contrast between passive
and intransitive, and of the way such distinctions are marked in the morphology and in the
syntax. The problem is exacerbated by a tendency in the past to transfer categories found in
Indo-European or Semitic to Sumerian.
Some linguists would say, for instance, that the contrast active - passive does not exist
in ergative languages. Karl Oberhuber, for example, in his examination of the Sumerian
passive, has said that: "Ein eigentliches 'Passivum' ist dem Sumerischen als einer
Ergativesprache von Haus aus Fremd" (1982: 133). Earlier, Diakonoff said that ergative
languages "have no grammatical direct object, from which follows that ( 1 ) no Accusative
can exist; (2) no Passive and Active voice can exist" ( 1965 : 1 8). However, Sumerian is
only split ergative; this means that such theoretical constraints may not apply equally to the
maru and to the gamt!! (although not everyone agrees about the degree to which Sumerian
198 Lesson 1 7
is split-ergative).
The question also hinges on the nature of the Sumerian verbal root. Most
Sumerologists now say that the Sumerian root is unmarked for voice or transitivity; that is,
du can be active or passive, transitive or intransitive. These categories are not marked at all
in the root; rather, they are determined by the syntax and semantics of the entire sentence
within which they occur. The parallel has been made with English sentences of the type
"he is cooking". This sentence, on the surface, can either mean "the man is cooking-up",
because of the temperature, or "the man is cooking some food".
In any case, it seems that the two constructions which are differentiated in English as
"intransitive" and "passive" are expressed by one construction in Sumerian. (This is one
reason why some Sumerologists say that Sumerian has no passive.) Therefore, some
Sumerologists use the compound term "intransitive-passive" (or "passive-intransitive") to
refer to both constructions. Other Sumerologists use the terms "passive" or "intransitive",
based on how the corresponding construction in English (or German) comes out.
The difference between transitive and intransitive verbs can be illustrated using du, "to
build", and gin, "to go". In the bamty, these are:
(1) The king went.
luga1.0 i.gin.0
(2) The king built the house.
lugal.e e.0 mu.n.du.0
In the intransitive sentence, the subject of the intransitive verb (the patient) is marked by .0.
This is cross-referenced by the .0 at the end of the verb. To express "the house was built",
which in English would be called a passive, Sumerian uses a construction identical with
sentence ( 1 ):
(3) The house was built.
e.0 Ldu.0
The .0 case-marker of the subject of the passive verb is cross-referenced by the .0 at the
end of the verb.
Sentence (3) is essentially the construction seen in line 17, although with the negative:
(4) gipar.bi.0 nu.(L)du.0
nu is the same negative marker seen in Text 1 6. There, it occurred before a nominal
form (an active participle). Here it occurs before a verbal form. As stated in Lesson 1 6, nu
fits into the category of modal-prefix. As such, it is regularly followed by one of the
conjugation-prefixes. Here the conjugation-prefix (.i ) has assimilated into the lul of nu.
The enclitic copula has occurred several times, e.g., in the PN dBa- -nin-am, "Baba
is queen". In addition to being used in such equational sentences, am can also be used to
express circumstantial clauses. These can usually be translated into English as: "it being
that", "it being the case that", or by a participial phrase in -ing. When am is used in such a
construction, it must follow a nominalized sentence. Hence the verb form is to be
understood as: [nu.(L)du.0].a.am, meaning "it being the case that its giparu had not (yet)
been built".
To sum up, am is the enclitic copula, used here to express a circumstantial clause. .a
nominalizes the preceding sentence. The underlying sentence which has been nominalized
Lesson 1 7 199
this is the analysis followed here. In Text 17, for example, there occurs: im-da-ab-sud4 -
re6. Under this assumption, this would derive from: Lb.da.b.sud.e.0. .i is the con
jugation-prefix, and .b is the optional pronominal-prefix which appears before the
comitative dimensional-prefix; here the reference is loosely ud, which is inanimate. The
phonological change may have been along the lines of I ibdal ) I iddal ) I imda/.
This explanation of im is plausible, because it is indeed normally followed by da or ta.
Lesson 1 7 201
many Sumerologists assume that the two conjugation-prefixes im-mi and im-ma somehow
derive from bi and ba (although it is not always made clear if this derivation is to be
understood in synchronic or diachronic terms). Falkenstein, for example, who has a
different understanding of the conjugation-prefixes than that presented in this book, derives
im-mi from *i-bi. Others, however, have argued for the existence of a morpheme fmf in all
these forms. It would roughly correspond in function to the Akkadian ventive (which is
normally used with verbs of motion, marking movement towards the speaker).
Such historical developments are unproven, and perhaps unprovable; Falkenstein, for
example, had to posit a number of apparently unmotivated phonetic changes to get his
forms to work. In general, some Sumerologists (such as Falkenstein) are inclined to see
variation in Sumerian as due to phonetic reasons, even if the rules governing the phonetic
changes cannot be determined; others (such as Jacobsen) are inclined to believe that
Sumerian grammar is more complex than usually believed, and that the variation we see is
due to our ignorance of the morphology, not to unexplained phonetic accidents. In any
case, in synchronic terms im-ma and im-mi (and im) pattern the same as the other conju
gation-prefixes; that is, they all occupy the same position in the verbal prefix chain, and
their presence is mutually exclusive. More work remains to be done on the synchronic
distribution of these conjugation-prefixes. Krecher (1985) is the latest attempt to isolate the
morphology and the semantics of the various conjugation-prefixes in fmf. He posits a
rather wide range of functions, which need to be more fully investigated.
- Case-markers
The presence of the two ergative case-markers (lines 14 and 20) has parallels in other
agglutinative languages. If a construction starts to become very long or convoluted, the
speaker (or writer) will occasionally "get lost" in the construction, and may occasionally
back-track, changing the topic, and will have to repeat a previous case-marker.
- History
The giparu at Ur was the "official" dwelling-place of the en-priestess (who is some
times referred to by the Akkadianized term "en tu-priestess"). It was a large structre,
composed of many rooms. The first such structure at Ur may go back to Early Dynastic
times; it was built and rebuilt right through the Neo-Babylonian period.
The en-priestess was always of royal blood. Perhaps the most famous was
Enheduanna, the daughter of Sargon of Akkad. She is well-known as the author of two
well-preserved poems, written in good Sumerian. In the Neo-Babylonian period, Nabo
nidus installed his daughter in the position.
The en-priestess "represented" the goddess Ningal, in some way. In particular, she
represented the goddess Ningal, while the reigning monarch represented the god Nanna
(the husband of Ningal) in some kind of "divine marriage ceremony". This ritual may have
originated as an end-of-harvest-time festival. It has often been discussed among Sumero
logists, and there has been much disagreement about what the sacred marriage was meant to
represent.
Penelope Weadock has summarized the functions of the giparu:
204 Lesson 1 7
Three separate units emerge from the Ur III--Isin-Larsa giparu building: the
Ningal temple which is the locale in which the entu-priestess, as the incar
nation of the goddess Ningal, carried out her most important function as a
participant in the rite of the sacred marriage; the giparu proper which was the
official dwelling of the entu-priestess, with its annexe, the cemetery for the
former entus; and the sanctuary in which the entu prayed for the life of the
king, her father or brother, in the hope that the gods would bestow prosperity
upon the land through the king, their human regent. (1975: 124).
It has been argued that the "institution" of the giparu existed in other cities, for ex
ample, in Uruk. This is probably true, but the evidence is not conclusive. However, Tex1
17 evidently refers to a giparu in Karzida, not to the giparu in Ur. This implies a Nanna
temple in Gaesh. Nothing is known of this temple or this giparu.
The original meaning of the term -par is unknown; it may have been "storehouse'
of some kind.
Lesson 1 8
This is a door socket of Shu-Sin, the brother and successor of Amar-Sin. He ruled
from 2037 to 2029 BC.
Sign-list and vocabulary
Notes
An-nu-ni-tum This may originally have been an epithet of Inanna. In Sargonic texts, the
d
divine name lnanna-An-nu-ni-tum is occasionally attested. But after the Old Akkadian
period, this compound term does not occur, only the individual term An-nu-ni-tum. Ac
cording to Jack Roberts, this pattern of attestation "suggests that the epithet split off and
became an independent deity" ( 1972: 147).
Gelb has pointed out "the tremendous number of compound divine names in the Ur III
period" ( 1987: 125). The name dlnanna-An-nu-ni-tum belongs to the class of names cha
racterized by Gelb as "DN plus description".
The etymology is unknown. It is possible that the -itum ending is an Akkadian femi
nine gentilic: II-t-um/. However, the meaning of "anun" or "annun" is unknown; it is not
sure ifit is Akkadian or Sumerian.
S u-dZuen The name is Akkadian, meaning "the one of Sin", or "the one belonging to
Sin". It was formerly read as "Gimil-Sin".
205
206 Lesson 18
Text 1 8
Notes
Many objects containing inscriptions have been damaged, either in ancient times or in
modern times. This means that part of the inscription may be completely broken away, as
in the bottom right-hand corner of Text 17, or part of it may be effaced, as is the top left
hand corner, and as in case 8.
It is important that transliterations reflect such damage, especially when the text is
completely broken. Unfortunately, there is no one system in use by Sumerologists (or
Assyriologists) to indicate such breaks. The most common system is to use brackets.
These are used to include signs completely broken away. Thus, line 1 1 is best trans
literated: --[ni]. Brackets can also be used to indicate partially broken signs. For
example, line 8 can be transliterated: lugal-[kala]g-g, and the last line as: mu-n[-du] .
Brackets are imprecise, however, in such cases as the partially effaced An-sign in line
. To resolve this problem, some scholars (not all) use half-brackets to indicate partially
broken signs. Thus, the first sign of line 1 can be transliterated rAn' .
Instead of half-brackets, some scholars use dots under vowels, to indicate that the sign
is partially broken or effaced. Thus, this An-sign might be transliterated as An. Some
scholars use different combinations of full brackets, half-brackets, and dots, in sometimes
rather idiosyncratic combinations.
The decision about whether or not to use brackets (or half-brackets) is not always
clear-cut. For example, what about the turn-sign in line 1 , or the first part of the Urims-sign
in line 9? In practice, such damage to the text is often ignored, if the context and the
remaining traces of the sign make the sign unambiguous.
Similarly, it is difficult to decide how breaks should be reflected in translation. Some
Sumerologists use brackets in translation, reflecting the breaks in the text. However, since
English and Sumerian are of such different grammatical structures (particularly in word
order), this procedure can be cumbersome and tiresome. Many scholars, therefore, omit
brackets i translation, especially when writing for a professional audience, because such
an audience will be able to follow or control the transliteration or autograph.
Occasionally, scholars may be suspicious of a published transliteration, or even an
autograph, of a cuneiform text. Or, they may wish further information about a partially
effaced sign. In such cases, they may collate the text, that is, physically examine the
cuneiform document. If the text is not easily accessible, they may ask another scholar to
perform such a collation.
The system used here is the simplest. Full brackets are used only to indicate
significant breaks. They are omitted from transcription and from translation.
Transliteration Transcription Translation
Discussion: structure
The structure of this text is:
[Annunitum, dam.ani].r benefactive
[Susin, ki.aga.a.Enlil.a(k), agent
lugal Enlil.e ki.aga.0.ag.anLa tn.pad,
lugal.kalaga, lugal. Urim.a(k),
lugal.anub.da.(k) limmu.bLak].e
[e.ani].0 patient
mu.n.(n.)du.0 verb
It thus follows the pattern of most royal inscriptions. This basic pattern, however, is
somewhat difficult to recognize, because of the length and the complexity of the
appositional phrases in lines 4 through 1 0.
-Relative clauses
The interpretation of lines 5-7 given above follows Jacobsen, who translates this
formulaic phrase as "the king whom Enlil envisaged in his loving heart". He considers it to
be a clause nominalized "in zero"; however, he cannot find many close parallels. These
lines illustrate a problem encountered several times before: a construction occurring fre
quently, its meaning relatively transparent, but its syntax dubious. There are still several
problems in the understanding of Sumerian relative clauses.
- Adjectives
There are a few cases in Sumerian where adjectives (or other modifiers) precede their
head noun, instead of following it. For example, the adjective kug, meaning "pure", quite
frequently precedes the names of gods and goddesses: kug-dInanna, "pure Inanna". It ap
pears to be the only adjective to be used so regularly in this position.
Other languages whose order is basically noun-adjective, such as French, also permit a
certain number of cases of adjective-noun constructions. These cases are usually limited to
a fixed number of adjectives or expressions. In general, it seems that languages of the
noun-adjective type permit more exceptions than do languages ofthe adjective-noun type.
- History
The circumstances of Amar-Sin's death are unclear. A late omen text says that he died
of an infection caused by a foot bite of some kind. Shu-Sin was his son and successor
(although it has been said that he was his brother, not his son). It was during the latter's
reign that trouble began to be felt in the empire: the Sumerians began to feel the pressure of
the Amorites.
Curiously, a number of "love poems" (to use Jacobsen' s term) have been preserved,
most of which are directed to the fourth king of the dynasty, Shu-Suen. One
guesses that this king, or perhaps more likely his queen, had in his entourage
a woman poet who enjoyed singing about love and lovemaking, and whose
210 Lesson 18
works, since they were cast in the fonn of praise for the king's beauty and
virile prowess, were favorably received and carefully preserved in writing
(1987:85).
Text 1 8 a
supplementary
This is a votive bead of carnelian, found at Susa. It has been speculated that this bead
was carried off to Susa as booty by the Elamites when they sacked Ur in 2004 BC, putting
an end to the Ur III Dynasty.
* llfl 1 1 \\
W $_ i fl1 l > l=I \-
Lesson 1 9
This door socket of Shu-Sin records one of the more significant events of his reign.
Sign-list and vocabulary
Nin-lil
- -
Ninlil (DN, fern)
u hand
ne forces, troops
gal great
Notes
Sar God of the city of Umma; son of Inanna. Not much is known of him; he did not
rank very high in the Sumerian pantheon.
Nin-l1l According to Sumerian mythology, Ninlil was raped by Enlil, and later became his
wife (darn). Nanna was their child. She was especially worshipped in Nippur.
21 1
212 Lesson 19
Mar-tu This term was used by the Sumerians in two ways. In an "ethnic" sense, it refers
to the (West-Semitic) Amorites, who dwelt in the Syrian desert west of Mesopotamia. It is
also used in a rather vague geographical sense, meaning "west" in general.
The etymology is unknown. Some now read it as Mar-du.
Mu-ri-iq Ti-id-ni-im The name is Akkadian, Muriq Tidnim, meaning "that which keeps
Tidnum away". muriq is the D-stem active participle in the construct state, from req!!, "to
be far away". Tidnum (or "Didnum") is probably the name of a particular Amorite tribe,
although here it refers to the Amorites in general. In late lexical texts, "Tidnum" is equated
with the Akkadian word used for the Amorites: ri-id-nu !!-mur-ru-y. The original form
=
and etymology of the name are unsure. In the various Semitic languages, several variant
spellings of the name Tidnum occur: Ti-da-nu-um and Di-ta-nu-um in Akkadian, Ddn and
Ttn in Ugaritic, etc. It is not always sure if such terms refer to the same people.
E-ag4-ge-pAd-da The principal temple of Shara in Umma. The name means "the temple
chosen in (his) heart", that is, e.ag.e.pad.a. The .e is the mark of the locative-terminative
case.
nir-g8.l It is difficult to determine the precise meaning of such a word; it is conventionally
translated into English as "prince" or "hero". The Akkadian equivalent is etellu, glossed by
the CAD as: "prince, lord".
In origin, this is probably an active participle with an incorporated object. gal normally
means "to be", but can also mean "to have". The combination nir-gal would mean "the one
who has nir". Unfortunately, it is not known what nir means.
ad-da Sumerian has three words for father: !!-!!,ab-ba, and ad-da. Lambert (1957) stu
died the distribution of the three terms on (primarily) geographical and class lines, but could
not come to any definite conclusions about their usage.
iib This is a very old loan-word into Sumerian from Akkadian waSipu. This Akkadian
term is usually translated "exorcist". iSib was then loaned back into Akkadian as iippu,
glossed by the CAD as "purification priest". iib itself is often translated as "incantation
priest".
It is never easy to determine the exact function of any priestly office. Iohannes Renger
(1967f) has studied the Old Babylonian priesthood, and exhaustively discussed the
Akkadian words for the different kinds of priests.
gudug The equivalent Akkadian priest is the paSgu, from the root paSau, meaning "to
anoint". The gudug-priest seems to be especially involved in certain kinds of ritual
activities, which it is not yet possible to define exactly. He was of a lower rank then the
iib-priest (at last in the Old Babylonian period); temples could have several gudug-priests,
but apparently no more than one iSib-priest.
This word is often transliterated as guda4 ' with the second vowel as lal, not as lu/.
sipad The cuneiform sign used to represent this word is actually a combination of two
Lesson 19 21 3
signs: the Qf!-sign ( * ) and the udu-sign <,ffi) . The Qf!-sign was originally a picture of a
staff or sceptre of some kind; in this meaning, it is read as gidri. It was also used to
represent the word for "overseer", Le., a man holding a staff of authority; in this meaning, it
is read as !!gula. The udu-sign means "sheep". Thus, the sipad-sign graphically represents
approximately "the overseer in charge of sheep".
It is reasonably sure that this word has a Id/-Auslaut. However, it is much more
common to find it transliterated as sipa.
ma-da This is usually assumed to be a very early loan into Sumerian from Akkadian matu
(although the latter is itself of dubious etymology; the only other Semitic language it occurs
in is Aramaic). The three terms kalam, ma-da, and kur are often translated into English as
"land", but they are not synonymous. Limet (1978) has studied their distribution, espe
cially in documents from the Ur III period. kalam is used exclusively to refer to Sumer.
kur originally meant "mountain". It then came to mean "foreign land"; it is never used to
refer to the land of Sumer. ma-da is more problematical; the sense is approximately
"territory". It is used mostly for foreign lands, but in certain uses it can refer to Sumer.
Jacobsen believes that in contexts such as Text 19 ma-da means "steppe". The opposition
between kalam and kur has also been studied by Steiner (1978) on a number of levels:
historical, legal, etc.
Because the term kalam is restricted to the land of Sumer, Poebel has speculated that
the word Ki-en-gi was a dialectal form of kalam. ka-na-.M is used as the Emesal equivalent
of both Ki-en-gi and kalam. However, in earlier texts kalam may have had a more general
meaning. Raphael Kutscher says that "Although the literal meaning of kalam is 'country'
(matum), it narrowed its scope to 'The Country' par excellence, namely, Sumer, and even
tually, to 'the nation', Le., the Sumerians" (1975:68).
dadag The reading is not certain; it is variously read as: da7 -dag, dag-dag, zalag-zalag,
zalzalag, and babbar2 ' The reading as I dadagl seems established by late lexical texts,
which give the syllabic writing da-da-f!g as the equivalent of the Akkadian word for "pure",
ebbu.
The dadag-sign is formed by the writing of two <) signs; the situation is similar to
that of the siskur1 -sign. The word Idadagl is probably a reduction of something like
*/dagdag/, a reduplicated form. The reduplication may have had some kind of intensive
value in its origin. This explains why there are two signs. Originally, one sign stood for
Idag/; two signs stood for Idagdag/. Later, */dagdagl was reduced to Idadag/, but the
original two signs continued to be written. Similarly, it has been speculated that the word
Isiskurl was originally a reduplicated form, perhaps */sikur-sikur/; this explains why it
was written with two signs. One would then have to assume a development along the lines
of */sikursikur/ ) */sisikur/ ) Isiskur/, or something similar.
2 14 Lesson 19
Text 1 9
15
BRIt>
o t> R est m istake ofscribe.
Lesson 19 21 5
Notes
Line 23: As the editors point out, the scribe has drawn a ni-sign, instead of the expected
du-sign. He may have been unconsciously influenced by the ni-sign directly above, in line
22.
Transliteration Transcription
1: dSara2 Sara
-
2: nir-Ml-An-na
- - -
nirgal.An.a(k)
3: dumu-ki-g2 dumu.kLaga.a
4: dlnanna Inanna.(k)
5: ad-da-ni-ir adda.anLr
6: dSu- dZuen Susin
7: mb-An-na iib.An.a(k)
8: gudug-u-dadag gudug.u.dadag
9: dEn-HI Enlil
1 0: dNin-lil-ka Ninlil.(a)k.a(k)
1 1 : .!! dig.ir-gal-gal--ne u digir.gal.gal.ene.(k)
12: lugal dEn-lil-Ie lugal Enlil.e
1 3 : ki-g2 kLaga.0
14: ag4 --na ag.anLa
15: in-pad Ln.pad
1 6: sipad-kalam-ma-e sipad.kalam.a(k). e
17: lugal-kalag-g lugal.kalaga
Ski-ma
1 8: -I-ygI-Unm lugal.Urim.a(k)
19: lugal-an-ub-da-limmuZ-ba-ke4 lugal.anub.da.(k) limmu.bLak.e
20: ud bad-Mar-tu ud bad.Martu.(k)
2 1 : Mu-n-iq Muriq
22: Ti-id-ni-im Tidnim.0
23: mu-du- mu.(n.)du.0.a.a
24: .!! ne-Mar-tu u ne.Martu.(k.)0
25: ma-da-ne- mada.anLe
26: bi-in-g4- bLn.gL0.a.a
27: E-ag4-g-pad-da Eagepada
216 Lesson 19
1: For Shara,
2: the prince of An,
3: the beloved son of Inanna,
5: his father -
6: Shu-Sin,
7: the iib-priest of An,
8: the gudug-priest with the pure hands
9: ofEnlil and Ninlil
1 1 : and of the great gods,
1 2: the king whom Enlil selected in his loving heart
1 6: to be shepherd of the land,
17: the mighty king,
18: the king of Ur,
19: the king of the four quarters -
are explicitly mentioned several times in Akkadian texts. The sense of the sentence is
probably: "the gudug-priest with the pure hands, in the service of Enlil and Ninlil, and in
the service of the great gods".
9- 1 1 . Enlil and Ninlil are conjoined, without any conjunction. These two deities are set
off from the more vague "great gods" by the conjunction y.
The two plural formations of nouns seen so far are the morpheme .ene (for animate
nouns), and reduplication (for inanimate nouns). Another common plural formation, used
when a noun is followed by an adjective, is to reduplicate the adjective: digir-gal-gal. This
type of formation is most common with animate nouns, but a few examples occur with
inanimate nouns: inim-gal-gal, "the great words". In addition, it is possible to tack on the
plural marker .ene: digir.gal.gal.ene, which is the form in Text 19.
Plural formations such as digir-gal-gal and digir-gal-gal-:-ne are usually thought to be
superlatives of some type: "the most great gods". This is very hard to prove. It is
especially hard to determine if there is any difference between the reduplicated forms with
.ene and the reduplicated forms without it. Only a few adjectives seem to occur in such
plural formations; by far the most common is gal.
In Text 6, the plural "king of the gods" was written illgal-digir-re- with redup
lication of the final Irl of digir. Line 1 1 of Text 19 uses a more morphemic writing; this is
the regular writing of this particular expression.
1 6. One of the functions of the terminative case in -e is to express the purpose of certain
verbs. Here the sense is: "selected to be the shepherd of the land". Expressions of this
type are not infrequent in the royal inscriptions. In them, the nominal phrase marked in -e
regularly follows the verbal form in-pad, instead of preceding it.
This deviation from standard Sumerian syntax is presumably to give some degree of
emphasis to this last constituent. There are instances where adverbial phrases, and even
patients, are placed after the verb phrase, instead of before it.
19. The .e marks the end of the ergative agent phrase, spanning lines 6 through 19.
20. Lines 20-26 form two when-clauses, governed by the ud of line 20:
20: when he built ...
24: and when he drove back ...
30: (then) he built ... .
The two when-clauses are linked by the conjunction y in line 24. The syntax of these
clauses is the same as that seen in Text 12. ud is the relative marker; it is followed by two
clauses, each of which is nominalized in .a, and each of which is in the locative case. A
more literal translation would thus be: "on ( the locative .a) the day that ( the norninalizer
= =
As for his great wall, it proved even more ineffectual than such barriers have
always been in the end. No more is heard of this vast and vain work, even
if, as seems likely, it furnished a line or a foundation for similar works in
later ages. Babylonia has no natural defences, and they were not to be pro
vided by an artificial rampart so long that it could have hardly been
effectually garrisoned ( 197 1 : 6 1 1).
It seems curious that this wall was given an Akkadian name, instead of a Sumerian
one. This may attest to the growing importance of Akkadian as the spoken language during
the Ur III period. Possibly it was chosen because its meaning would have been more-or
less understandable to Amorite speakers.
- History
The ninth year of Shu-Sin commemorates the building of the temple mentioned in this
inscription: mu _ dSaraTUmmaki-ka ba-du, "the year when the temple of Shara of Umma
was built".
Such year-dates, as laconic as they are, actually comprise one of the principal sources
of information about the history of the Ur III period.
Text 1 9a
supplementary
Commentary
arad servant
Notes
223
224 Lesson 20
According to Hallo, the term means "titular head of a city and its dependent territories".
Jacobsen says:
The title ensik ... seems to denote specifically the ruler of a single major city
with its surrounding lands and villages, whereas both "lord" (en) and "king"
(l!!gill) imply ruler over a region with more than one important city. As for
the origins of the office, the ensik seems to have been originally the leader of
the seasonal organization of the townspeople for work on the fields:
irrigation, ploughing, and sowing (1957:384 n.7 1).
But later, during the Ur lIT period,
The top provincial civil administrators, the ensiks, became proper governors,
entirely dependent on the king, and were moved at will from one post to
another to minimize the dangers inherent in too strong local ties. Military
affairs were out of their hands entirely (1957: 155).
During the Ur ITI period, the ensi2 was the highest-ranking civilian authority; the cor
responding military officer was the akkana (often translated "military governor"). Many
of the latter were sons of the king. Occasionally, one and the same person served as both
ensi2 and akkana. The names of many ensi2s of the Ur III period have been preserved.
ensi2 was loaned into Akkadian, appearing as ii)akku and iakku. It is glossed by
the CAD as: "territorial ruler (of cities, countries, etc)".
The etymology and writing of the word are discussed below.
arad The original form of this sign was the nitab-sign (meaning "man") followed by the
kur-sign (meaning "mountain"). Presumably, the Sumerians derived some of their slaves
from foreign, mountainous areas. The shape of the arad-sign, however, tends to become
simpler, and even as early as the Ur lIT texts, the nitab-sign and the arad-sign can look quite
similar.
The Akkadian word for "slave" is wardu. This would seem to derive from the verbal
root waradu, meaning "to go down to"; this root occurs in other Semitic languages. There
fore, most scholars assume that the Sumerian word arad was borrowed from the Akkadian
wardu. A minority view, however, says that wardu has nothing to do with the verbal root
waradu, and that wardu is a borrowing from Sumerian.
A complicating factor is the fact that the reading of the sign is uncertain; some read it
as arad, and others read it as ir. The Sumerian evidence is ambiguous. According to Gelb,
"generally, the form ending in -d is younger than the form ending in -r" (1982:86). Thus,
in pre-Sargonic Lagash there occurs: (NITA1jxKUR)-ra-ni, "his slave", but in Sargonic
and Ur III texts there occurs: (NITAIJxKUR)-da-ni, "his slave".
Most scholars, however, seem to derive the form in Irl from that in Id/. For
example, Falkenstein assumes a development along the lines of: */ward-al > */urd-a/ >
*/ird-a/ > */irdl > fir!. Gelb, on the other hand, believes that there were originally two dif
ferent words for slave: a native Sumerian word, written ITI I (ARADxKUR), and also a
borrowed word, written arad (NITA1jxKUR).
There has been much recent discussion about slavery in the Ancient Near East. The
term arad is variously translated as "servant" or "slave". But as Sollberger (and others)
Lesson 20 225
have pointed out, "The usual translation 'slave' is a misnomer because its legal implications
do not fit the Sumero-Akkadian social context" (1966: 137). In a similar vein, Gelb has
said, "Freedom is relative and the terms for 'slave' are quite ambiguous in the Ancient Near
East, as they are in the Classical World, or for that matter anywhere else" (1979a:284).
226 Lesson 20
Text 20
n I " - .----
c;:::;:;;-- 1111 r
Lesson 20 227
Notes
Line 5: The pad-sign is here split into two parts. This is a common scribal practice with
this sign. It is presumably done for aesthetic reasons; otherwise there would be too much
empty space in the line.
Transliteration Transcription Translation
to avoid a possible prejudging of the written form. The latter view would regard trans
literations of the type "arad-da-ne" as an attempt to jiggle the script to fit our views of
Sumerian grammar. A similar instance was seen in Text 19, where ma-da-ne- occurred;
others would read this as ma-da-ni-. It is also possible to interpret writings of the type
arad-da-ni- as historical or morpho-graphemic spellings. It is such ambiguities in the wri
ting system that make it difficult to establish rules governing Sumerian phonology, espe
cially those governing vocalic contraction.
Discussion: structure
The bare-bones structure of this text is:
1 -9 For Shu-Sin benefactive
1 0-13 Habaluge agent
14 his beloved temple patient
15 built. verb
In all the inscriptions in the body of texts seen up to this point, the initial datival phrase
referred to a god or goddess, the deity for whom something was done or built. However, it
is also possible to find inscriptions recording actions performed by a subordinate of some
kind, in order to curry favor with the king. Hallo considers these to be royal inscriptions,
because there is a fair amount of latitude in his definition of royal inscriptions: inscriptions
which are dedicated either "by, or to, or on behalf of the king".
There are not many such inscriptions. Hallo lists only five building inscriptions of this
type. Curiously, they all come from the reign of Shu-Sin. king.
- Phonology
There are other cases of Idl - Isl alternation in Sumerian. Sometimes this alternation
shows up in different syllabic writings of Sumerian, sometimes it shows up in differing
Akkadian versions of loan-words or of proper names. The goddess of the scribal art, for
instance, variously appears as Nidaba or Nisaba. Some Sumerologists think that this
alternation means that Sumerian had a voiced interdental fricative, IM, which it was
difficult for the script to represent.
There are several cases of bisyllabic (or polysyllabic) Sumerian words which show
vowels of only one quality, but whose Akkadian equivalent shows vowels of two different
qualities. For example, there is a tree usually spelled za-ba-Iam in Sumerian. The word
appears in Akkadian as supalu. Most probably, this is a pre-Sumerian substrate word,
which passed into Sumerian. It then passed into Akkadian, presumably through Sumerian
(or conceivably by a different route). The Akkadian preserves the older vocalization; the
Sumerian form shows that at some time Sumerian underwent a rather extensive process of
vocalic assimilation, a process of which only traces can be seen. Assuming that Akkadian
borrowed the word from Sumerian, it must have borrowed the word before the Sumerian
vocalic assimilation had taken place. This same process was seen in Lesson 1 1, where the
word zabar was discussed; the pronunciation Izabarl is the result of vocalic assimilation
from some earlier form of the word like Isibar/; Akkadian again preserves the more
Lesson 20 229
original vocalization.
- ensi2
There is no unanimity among scholars about the etymology and the writing of the term
ensi2. (The latest summary is in Dunham 1986:5 1-52.)
The word probably had a Ik/-Auslaut. When it is followed by a vocalic ending of
some kind, a Ikl usually appears; note also the Akkadian ii'akku. It has frequently been
proposed that the Ikl is the genitive marker; the word may originally have been a genitive
phrase, en. si. (k), "the lord of si". Unfortunately, it is not sure what si means here.
Jacobsen thinks that si means "arable land"; the title would originally have meant "manager
of the arable lands". But it has also been argued that the word has a pre-Sumerian substrate
etymology, and that the interpretation "lord of si" is a Sumerian folk-etymology.
Some Sumerologists now transliterate this word as enS -si. This interpretation is
partially based on the understanding of the word as a genitive construction, even though the
spelling with en S (instead of simple en) is weird.
The problem of understanding the writing is complicated by the fact that although
normally written PA-TE-SI, the word is also occasionally written PA-SI, and even just PA;
other bizarre spellings are attested. These writings make it difficult to understand how the
cuneiform signs are meant to refl ect the language (should they be regarded as
abbreviations?). Such writings cause numerous problems in transliteration.
-Genitive
One of the pieces of evidence which indicates that the final Ikl of the genitive
morpheme was actually pronounced in spoken Sumerian is the fact that it occasionally turns
up in words loaned into Akkadian. For instance, the Sumerian god Amar-Utu, "young bull
of Utu", appears in Akkadian as Marduk, presumably from amar.utu.(k) If the Sumerian
ensi2 is indeed a genitive phrase, then the Akkadian ii)akku also shows the genitive mar
ker Ik/. It has, however, also been argued that "Marduk" is not of Sumerian or of Ak
kadian origin, and has nothing to do with "Amar-Utu".
- Proper names
The personal name "Habaluge" occurring in this inscription is presumably of
Sumerian origin, although the exact etymology is not known. When a name is of uncertain
etymology, it is often difficult to determine the precise reading of the signs. For example,
the third sign in this name can be read Iu S ' lul, lab, nar, etc. The fourth sign is also gra
phically ambiguous; several different signs share approximately this same shape. For
example, the -sign and the ke4 -sign, even though distinct signs, look very similar in sev
eral of these royal inscriptions. And if it is the ke4 -sign, it has several different readings:
ke4' , lil, etc.
One way such ambiguities can be resolved is to identify different spellings of the same
name. For example, this particular name is not uncommon in Sumerian texts. It is most
commonly transliterated as: Ij-ba-lu S -ke4' with the proviso that the reading of the third
and fourth signs is not certain. There are also spellings where the third sign appears as lu,
230 Lesson 20
and not as the lus -Iul-Iab-nar-sign: lj-ba-Iu-g and lj-ba-Iu-g 18 . This shows that the
third sign is probably to be understood as Ilu/, and so to be read as lu, and not as lul, lab,
or nar. The fourth sign is spelled as , g, and g 18 : lj-ba-Ius -' lj-ba-Iu-, and lj
ba-Iu-g 1 8. This shows that the last sign is a g/ke4 -sign, and not the graphically similar
sign; it also shows that the reading was probably Ige/. Therefore, the most likely reading
of the name in Text 20 is: lj-ba-Ius-g.
The latest discussion of this name is by Steinkeller ( 1984:9); he reads the third sign as
lugx' that is, with a Ig/-Auslaut. He believes that the meaning of the name "cannot be
gauged with confidence", but it is undoubtedly a Sumerian verbal form with the modal
prefix b; it may mean "May-he-pasture/take-care-of '.
Lesson 2 1
The three texts in this Lesson do not offer very much new in the way of grammar, but
they illustrate common types of royal inscriptions. The first is a weight of Shu-Sin. The
next two are seals of rbbi-Sin, the son and successor of Shu-Sin; he was the last ruler of
the Ur m Dynasty.
W "5"
ma-na "mina" (measure of weight, about 505 grams)
Notes
5 When Sumerian numerals are used strictly for counting, they are normally transliterated
by Arabic numerals. The word for "five" was iA.
Text 2 1 a
231
232 Lesson 21
Discussion: weights
This is a typical weight inscription. Unlike the weight inscription seen in Lesson 10,
this one actually gives the weight. It is difficult to say exactly what gi-na means in such
contexts; it is usually understood as "standard", or perhaps "certified". One of the more
important functions of Mesopotamian rulers (throughout all periods of history) was the
regulation of the system of weights and measures, but not much is known about how such
weights were actually managed by the crown. Gadd points out that Shulgi "rearranged the
calendar, set up a bureau of standards, and issued accurate weights which were preserved
and imitated to the latest days of Babylonian history" (197 1 :618). In the prologue to Ur
Nammu's Law Code, there is a section referring to the "standardization" of the mina; this is
usually interpreted as referring to some kind of reform of the royal weight system. Re
cently, Irving Finkel has published a text dated to Amar-Sin's first year, which is a "receipt
for two differing sets of weights":
The implication of the text ... is that an official issue of correct weights was
made at the beginning of Amar-Sin's reign, and that this document reflects a
deliberate attempt to ensure that government offices were using uniform
weights (1987 : 192- 193).
- Metrology
All the Classical Semitic languages except Ethiopic have a verbal root *mnw/y,
meaning "to count, to reckon". Therefore, Sumerian ma-na is probably a loan from the
Akkadian manu (which is also the ultimate source for the English word "mina").
The value of the ma-na and the manu varied to some degree from time to time and
from place to place (Powell refers to "a multiplicity of standards which defies reduction to
one or more 'common' norms"). In Sumer proper, the most common value of the ma-na
was about 505 grams. In Mesopotamia, the manu was the same. But in most of Syria, the
Lesson 21 233
manu was a little less, from 470 to 480 grams. At Ebla, it was also about 470 grams. (The
theoretical value of weights is determined simply by averaging out the actual weights of
weights inscribed with their values.) Since this particular weight is a .5. ma-na weight, it
should weigh about 2525 grams. It actually weighs a little less, 25 1 1 grams.
The Sumerian ma-na was divided into 60 gin; the Akkadian manu into 60 iqlu. 60
ma-na formed a gun; 60 manu formed a biltu:
= 60 ma-na 1 biltu = 60 manu
= 60 gin 1 manu = 60 iqlu
- Numbers
The pronunciation of the numeral for "five" as fia/ is known from mathematical
cuneiform texts, where it is occasionally spelled out. The Ebla school-text mentioned in
Lesson 10, which spells out the Sumerian numerals from one to ten, simply gives i for "5".
Unfortunately, not enough is known yet about the nature of the Eblaite syllabary to say
what values the i-sign could have had at Ebla; however, there is some evidence to show that
one of its values was /ya/. Pettinato, in fact (198 1 : 143) reads the sign on TM.75.G.2198
as i, but this is perhaps somewhat adventuresome.
It is also significant that the i-sign is composed of five strokes: ;::i= .
>--
zu your
Notes
I-bi-dZuen The name is good Akkadian, meaning "Sin has called"; ibbi is the preterite
from nabu, "to call".
D a-da The etymology of the name is unsure; it is discussed further below.
Nanibgal Very little is known about this deity. The reading of the name is somewhat un
certain, as is the etymology. The cuneiform character appears to be the an-sign followed by
the nisaba-sign. Therefore, the name is sometimes transliterated as AN.NISABA, or as
234 Lesson 21
Text 2 1 b
Notes
The name of Dada's father, occurring in line 1 1 of this text, illustrates the difficulties of
working from autographs. The autograph seems to show the name as Ur-t!gS . t!gS (or
a6) means "good". The name would then be a variant of Ur-ags-g, a relatively common
personal name meaning "the good man". However, the photograph of the text is more
ambiguous. The sign in question is actually somewhat damaged, and the remaining traces
can be made to fit either ags or dNanibgill.
The reason for preferring the reading dNanibgal over t!gS is because other seal im
pressions have been preserved of this same individual, Dada, in which the sign for the
name of his father is more distinct, and in these the sign is clearly Nanibgal.
One cannot always accept a modern-day editor's transliteration of a text. An autograph
carries more evidential value, but even then cannot always be accepted at face value. This is
especially true for autographs which were drawn when knowledge of Sumerian was weak.
Better than a transliteration or an autograph is a photograph. However, for many published
texts no photographs are available, and for others the photographs are reproduced in such
poor quality that they are almost useless. For any significant passage, there is no alternative
to a close examination of the original cuneiform document.
Lesson 21 235
The initial consonant of the marker for the firstperson singular is the velar nasal
discussed under Phonology. conventionally transliterated by Sumeroiogists as g. There
fore, gylO is the most up-ta-date reading of this sign in this usage. However, it is also
transliterated as gylO' and (especially in older works) as mu.
- Proper names
The name Da-da occurs very frequently in cuneiform texts. It is not very easy to de
termine the etymology of a name of such a simple structure, what appears to be a redup
licated CV syllable. The name has variously been considered to be Sumerian, Akkadian, or
"other", In an article on "Ethnicity and Onomastics in Sargonic Mesopotamia" (1982),
Benjamin Foster divided personal names into four groups: Sumerian; Akkadian; Redup
licated; Unsure. He purposely omitted the name -IDt from discussion, because of the
difficulty in detennining its etymology.
Such reduplicated personal names are sometimes referred to by the Gennan tenn
"Lallnamen" or "LallwOrter" ; another example is Du-du.
-Seals
Seals were used by officials in Mesopotamia (and elsewhere) to stamp their "seal of
approval" upon documents of all kinds. The act of sealing could perfonn several functions,
such as acknowledgment, authorization, guarantee, etc. Mesopotamian seals usually
consist of two components: a pictorial scene of some kind, and a short inscription. Many
seals have only a pictorial scene, and lack an inscription. There are only a few seals which
have just an inscription, but lack a pictorial scene. The pictorial representation on the seal
frequently has a mythological significance, as in the example given below. Or, it may be a
"presentation scene", usually thought to represent the possessor of the seal paying homage
to his ruler, the king. The following is a picture of a (non-royal) seal, from the Old
Akkadian period. (This partiCUlar seal has often been reproduced.) The inscription reads:
M-d..!! mm-sar, "Adda the scribe". The photograph is actually of the impression of the seal,
not the seal itself. This is because the text on seals is inscribed in reverse (that is, mirror
imaged), so that when impressed, the text comes out in the right direction.
I II
'f ,
)j\; I
'
A
, \ '-'-'-'-
-
Lesson 21 237
It is not easy to detennine what language such a short inscription is written in. It may
be perfectly good Sumerian. However, according to the criteria used by art historians, the
seal in the photograph dates to the Old Akkadian period. Therefore, the language is
probably Akkadian, and dub-sar should be read as a logogram for tuparru.
- Scribes
Scribes occupied a central role in all of Mesopotamian civilization. Many modem
scholars have written about them, but there is still much that is not known; for example, to
what degree was their job hereditary; what was the extent of their influence at the royal
court; did they have non-scribal work at the court or elsewhere, etc. Writing on the "social
position of Neo-Babylonian scribes", Muhammad Dandameyev has said:
Our infonnation about the social position of the Mesopotamian scribe, his
activity as bureaucrat and in the service of the community for recording of
contracts is very scanty. We have no direct data on the economic situation
and the social origin of scribes. We also do not know if the scribal
profession was the chief source of income or if the scribes were busy with
their craft along with handicraft, tilling ofland and so on ( 1982:35).
C. B . F. Walker points out that
The scribes, like any craftsmen, had to undergo training, and having
completed their training and become entitled to call themselves dubsar
"scribe", they were members of a privileged elite who might look with
contempt on their fellow citizens (1987:33).
Specifically discussing seal-practice in the Ur III period, Steinkeller says:
The tenn dub-sar, apart from its basic meaning "scribe" is an honorific title
which merely indicates the graduation of the individual in question from a
scribal school. ... It is tempting to speculate that the "dub-sar seal" was a
kind of "diploma", which may have been presented to a graduate of a scribal
school at the conclusion of his studies. The possession of such a seal would
have constituted proof that its owner was eligible and entitled to be employed
in the state or temple administrative apparatus or to sell his services to private
individuals ( 1977:47-48).
In a similar vein, Veenhof mentions the role of scribes
in the administration and their position in society, which may range from
that of a simple clerk or a paid letter writer on the market to that of a chief
accountant or secretary of a chancery or king (1986:2 1 ).
The rather automatic translation of dub-sar as "scribe" paints a rather simplistic picture.
Michalowski says that "In Ur III times dub-sar was a general term for low and middle level
bureaucrats" (1987:62).
And to quote Walker again,
[Most scribes,] after all their technical training, spent their lives writing lists
of deliveries of sheep or issues of barley rations and occasionally taking a
letter by dictation. The more successful scribes would end up as senior
administrators in the state bureaucracy, but most of their colleagues would
238 Lesson 21
have been happy simply with their status as educated men and the
knowledge that their training guaranteed them employment ( 1987:39).
In a thought-provoking article about what we don 't know about Ur III society, Soll-
berger asks:
We know roughly what the professional scribe's jobs consisted of, but how
did he work, and where did he work, and how did he make himself known
as a professional scribe and his services available? And there is of course
the nagging question which is usually politely glossed over: where did the
scribes get the enormous amount of clay they needed? Were there clay
stationers? Did one have to buy clay or did one just go to the canal bank and
help oneself ... ? ( 1972: 1 88).
- Engravers
When dealing with "monumental" inscriptions, it is necessary to distinguish between
"scribe" and "engravers" (or, "lapidaries"). The latter were the persons who actually
chiselled the inscriptions into the stone. They were not always literate, but simply copied a
design or plan, which may have been drawn onto the stone. Presumably, the engraver
worked under the supervision of a scribe. In other cases, the scribe and the engraver may
have been the same person.
The standard word for "engraver" was zadim. This word apparently derives from za,
"stone", and dim, "to fashion"; dim is an active participle, and za is an incorporated direct
object.
In the case of seals, the situation was probably a little more complicated. The
inscription and the pictorial scene were sometimes engraved by different individuals. W.
G. Lambert (discussing seals of the Cassite period) asks:
A basic question which needs answering is, who carved these inscriptions?
Did one man carve both glyptic and inscription, or were separate craftsmen
employed for the artistic and scribal parts? In some cases it is clear that the
glyptic was carved first, because not enough room was left for the
inscription, so that the last line had to be spread out among the glyptic. But
in other cases where the inscription covers virtually the whole area, and the
glyptic is reduced to a row of insects for example, then one may suspect that
the inscription was carved first and the glyptic was a second thought, serving
merely in fugam vacui. ... One may wonder if two quite separate guilds of
craftsmen were in existence, and such a division of labour seems very
probable in the contemporary boundary stones. ... On general grounds too
such a differentiation is likely, since the artist and the scribe needed very
different training. Yet one need not suppose that this demarcation was al
ways completely enforced ( 1975:220).
- Seals
Many seals from the Ancient Near East have been preserved. Even more common
than the seals themselves are seal impressions, that is, the impression of a seal upon a
Lesson 21 239
za, presumably ARAD-sa, for Iwarassa/, "her servant", and another seal written geme2-za,
,
possibly GEME2-sa, for lamassa/, "her servant (fern) '.
While this may be true for seals of the Akkadian period, it is hard to say whether it
might be true for Ur III (and other) seals. There is really no evidence to decide one way or
the other, and so the question is still open. Barring explicit evidence to the contrary, it is
probably best to assume that for the Ur III period, at least, the text is Sumerian.
- Seals
Both Text 2 l b and 2 l c may be called functional, in the sense discussed in Lesson 4.
There are also a certain number of votive seals. These are known primarily from the seals
themselves; only a few of the actual impressions are preserved. Text 22, below, is such a
votive cylinder seal.
sa (kind of priest)
Notes
Sag-dNanna-zu sag means "head", but can also have the meaning "slave". zu is a verbal
root meaning "to know". Names of this type are usually interpreted as some kind of
reduced relative or participial clause, "the slave who knows Nanna", or "the slave whom
Nanna knows".
saga This priest was high up in the temple hierarchy, although very little is known of his
priestly duties. He seems to have been mostly concerned with running the administrative
side ofthe temple. While the conventional translation is "sang-priest", Gelb points out that
"In his capacity as the head of a household, the word san may be interpreted as the chief
administrator of a temple household" (1979b: 1 6); Snell translates the title as "economic
director of a temple". The same cuneiform sign, in fact, can be read as ita5 ' "to count"
(Akkadian manu); Aage Westenholz, among others, says that the sign "depicts an abacus or
a countingboard" ( 1985:296).
Lesson 21 24 1
Text 2 1 c
- History
Shu-Sin apparently died a natural death. He was succeeded by his son, Ibbi-Sin.
Early in the latter's reign, the eastern territories under the control of Ur broke away, then
other parts of the empire began to fall away. For most of his reign, Ibbi-Sin's control
extended no further than the city of Ur itself. The economy collapsed, and a vicious
inflationary spiral ensued. Very little is known of the details of the twenty-four or twenty
five years of Ibbi-Sin's "reign". Jacobsen has said:
How an empire like that of the Third Dynasty of Ur - to judge by our
sources the most efficiently organized structure of its kind before Assyrian
times - could so quickly and so completely collapse without pressure from
any enemy state or states of comparable magnitude is really quite puzzling
(1953: 173).
Jacobsen wrote this over thirty years ago, but his puzzlement still largely stands;
Steinkeller says that "the phenomenal rise of this empire was matched only by the sudden
ness and completeness of its demise; in less than a century after its creation, no trace of it
remained" ( 1987b: 19).
Finally, there came invasions by the Amorites, against whom Shu-Sin had built the
wall mentioned in Text 19, and the Elamites. The Elamites, aided by a somewhat obscure
group of people from the Zagros mountains known as the "Su" or "Sua" (recently
identified by Steinkeller [1988] with Shimaski in Iran) sacked Ur, then withdrew back to
Elam, carrying Ibbi-Sin back with them; he died in Anshan. Gadd says: "Ibbi-Sin became
the typical figure of an ill-starred king, remembered only for his captivity and death in a
strange land" (197 1 :6 17).
During their sack, the Elamites destroyed every temple standing in Ur, and all of its
fortifications. One of the most well-known Sumerian literary compositions is a long poem
entitled "The Lament over the Destruction of Ur", which bemoans its destruction.
This was not the only time that Ur was sacked; Samsu-Iluna of Babylon also levelled
the city, in 1740 BC. The year-date for the eleventh year of his rule is: "The year in which,
at the pleasure of Anu and Enlil, he destroyed the walls of Ur and Uruk". Woolley
describes the destruction:
The ruins bear eloquent testimony to the thoroughness of that destruction.
The fortifications were dismantled - this indeed one might expect; every
temple that we found had been plundered, cast down, and burned; every
house had been consumed with fire; the whole of the great city ceased to
exist (1982:214).
However, the city was rebuilt almost as often as it was sacked. Most kings of the
Isin-Larsa period rebuilt old temples and built new ones. Such construction took place
right through the Neo-Babylonian period. Although Ur never regained the political
importance it enjoyed under the Ur III Dynasty, there were times when it must still have
been an imposing city.
Ur was occupied - at least, to some degree - into the Persian period. It began to
completely fade out about the fourth century BC, through the effects of changing trade
Lesson 21 243
patterns. the shifting of the course of the Euphrates and the concomitant loss of agriculture,
etc.
Text 2 1 d
supplementary
Notes
Line 8, In Text 21a, gi-na (a borrowing from Akkadian) was used as an adjective.
However. it can also be used as a verb: mU-!li!-gi-in. It is usually translated as "he stan
dardized". or perhaps here "he certified". As discussed above, the exact significance of the
tenn is not clear.
The actual weight of this weight is 2478 grams.
Lesson 22
This is a votive cylinder seal. made of limestone. dedicated to the tife of Shulgi.
245
246 Lesson 22
il
Notes
Mes-lam-ta-e-a God of the Netherworld, apparently the same as Nergal. Mes-lam (also
transliterated as MM-lam) is the name of the temple of Nergal in Kutha. Its meaning is
unknown (although gimes is well known as a type of tree). ta is the ablative case-marker.
r:) )I:::j is a verbal root, meaning "to go out". is the nominalizer. Thus, the name means
something like "he who goes out of the Meslam-temple".
Meslamtaea himself does not occur frequently in texts, but Nergal was widely
worshipped in Mesopotamia. He shared the rule of the Netherworld with Queen
Ereshkigal.
Ki-lul-la This not uncommon name is presumably Sumerian, although the etymology is
unsure. It is variously spelled as: Ki-lul-la, Ki-lul-hi, and Ki-lul-hl-.
Ur-ba-? The reading of the third sign is uncertain. Both the autograph and the photo
show a sign which looks closely like the last sign in the inscription, which must be bi. The
reading bi for this sign is also the reading preferred by Gadd, who says that upon a "fresh
examination" of the seal, "the engraver certainly traced, and doubtless intended, the same
sign as in the last place of the whole inscription", that is, bi. However, there seem to be no
parallels to a putative name Ur-ba-bi. (Gadd speculates that Ba-bi may be a variant of the
name of the goddess seen in Lesson 1 1 , spelled as dBa-.!:2% , which earlier Sumerologists
read as dBa-.!1.)
Most Sumerologists have interpreted the sign as a poorly drawn -sign (if not a
simple mistake). The g-sign here would be read as gar. Ba-gar is the name of a well
attested temple of Ningirsu at Lagash (the etymology of Ba-gar is unsure). The personal
name would then mean, "the man of (the temple) Bagara". This name is attested elsewhere.
Also, other names composed of Ur and the name of a temple are attested.
Lesson 22 24 7
Lagd "Lagash" is used in two senses in Sumerian (and in English). Strictly speaking, it
refers to a city proper (whose modern name is el-Hibbe). But it can also be used to refer to
the territory controlled from the city of Lagash. This larger entity is sometimes referred to
as "Lagash-state", as opposed to "Lagash-city". This included Lagash-city; the adjacent
Girsu (whose modern Arabic name is Telloh, meaning "mound of the tablet"), and Nina
(modern Surghul). Because of this ambiguity, the term "Lagash" is sometimes used when
referring to Girsu.
This was the site of the first important excavation of a Sumerian tell, begun by the
French in 1 877. Thousands of tablets were found, including a number of royal inscriptions
from the First Dynasty of Lagash, and from the time of Gudea. These are one of the prime
sources of Sumerian for the period.
The etymology of the name is unknown. It is not known how the three signs SIR
BUR-LA came to represent the name of this city. One would guess that the la-component
is some kind of phonetic complement. The name is, in fact, occasionally written LA-BUR
SIR.
It This originally meant " arm", and then "strength". In this latter sense, its Akkadian
equivalent is eml1q!!, glossed by the CAD as: " 1 . strength (in physical sense as localized in
the arm), . .. ". J!-zid-da is something like "the effective arm". It can be approximately
rendered in English as "right arm" or "right-hand-man".
In Text 17, the expression digi!:-zid occurred, translated as "the effective god". There,
zid appeared without the nominalizer -. But here, the same adjective occurs with the
nominalizer -: zid-da.
gu-za-llt g!!-za means "throne". hi is a verbal root meaning "to hold, to lift, to carry".
Here, hi is an active participle, and g!!-za is its incorporated object: "he who holds the
throne", or "throne-holder".
gg-za-la was borrowed into Akkadian as guzalU, explained by the CAD as: "an offi
cial, lit. chair-bearer, originally a servant carrying a chair after his master".
gg-za is thought to be the ultimate source of the word for "chair, throne" appearing in
many Semitic languages: Akkadian kussu, Hebrew kisse, Arabic kursTy, etc. But since the
pattern of the Sumerian form is a little strange, it has more recently been speculated that gg
za is a borrowing from Akkadian. This situation illustrates the difficulty of evaluating the
evidence of loan words; it is not always easy to establish the direction of a borrowing.
kBib It has been speculated that this is a pre-Sumerian substrate word.
getug3 This can be spelled in a variety of ways. The simplest was originally the picture
of a donkey's ear ( <T>-), transliterated as getug. It can also be written with the signs
and illg functioning as phonetic complements; the combination of these three individual
signs can appear in varying orders:
ge-tuggetug ( getug2)
=
gegetugtug ( getug3 )
=
Although it is clear (now) that ge and illg in this particular use are phonetic
248 Lesson 22
complements, they are not usually transliterated as such. Since the simplest (at least,
phonetically) reading of the getug-sign is m, Sumerologists often refer to this sign as the
"m-sign", and transliterate this word as: ge-tUg-m, or (as here): ge-ill-tug.
In its concrete meaning as "ear", the Akkadian equivalent of getug is uznu. However,
it is also equated with basTsu, glossed by the CAD as: "(1) aperture of the ear, ear, (2) (fa
culty of) hearing, (3) understanding". Jacobsen says that "the Mesopotarnians believed the
ear, not the brain, to be the seat of intelligence".
ags Variously transliterated as: ag5 ' a6 ' sag9 ' sa6 ' sig6 , etc. The most common Akka
dian equivalent is damliqy. The verbal adjective damqy is glossed by the CAD as: " 1 .
good, fine, pleasant, 2. beautiful, handsome, 3 . of good family, well-to-do, 4 . expert,
well-trained, 5. of good quality, in good condition, 6. gracious, favorable, 7. propitious,
8. effective, 9. canonical".
Lesson 22 249
Text 22
* J:1>J C} E }F
j3 H .El
-+ Efl tIr ni l I
, (
t ::\r ,
* JIIUt, ,::><J
u
l
250 Lesson 22
Notes
Line 9. Assyriologists use a superscript exclamation mark ( ! ) to indicate that the cuneiform
sign so marked in transliteration is written in a deviant or aberrant fashion on the tablet.
For example, the proper name in this text might be transliterated as: Ur-ba-gar ! . This
transliteration means that in the opinion of the modern editor, the sign is a poorly drawn
gar-sign.
Another convention is to employ the superscript exclamation point to mark the correct
value in transliteration (that is, what the editor thinks it should be), and to follow this with
"what is actually written", in caps within parentheses. Thus, a transliteration such as: Ur
ba-gar! (BI) would mean that in the opinion of the editor, the sign drawn on the tablet is
bi, but the editor believes that this bi-sign is a mistake for gar - the scribe made an error.
As can be imagined, these conventions can be misleading or confusing. They illustrate
the importance of working directly from the texts, from photographs when possible, or
from autographs, and not just from transliterations.
Line 10. Because of damage to the cylinder seal, this line is somewhat difficult to read. Of
the getug3-sign, the ge-component o;=f) is clear. The ID (or getug)-component J>--) is
virtually completely effaced. Of the !!1g-component ( j' ), only the bottom shows. In
addition, the following nig-sign ( ilY") is scrunched up.
Transliteration Transcription Translation
1 : des-lam-ta--g eslamtaea For eslamtaea,
2: lugal -zid-da lugal a.zid.a the king, the right-hand man
of
3: Lawki -ke4 Laga.(a)k.e Lagash -
4: nam-.ti-il nam.til for the life of
5: dSul-gi nitab-kalag-gg Sulgi nitab.kalaga Shulgi, the mighty man,
6: lugill-Urim5ki-ma-ka-e
- -- -- - - lugal.Urim.ak.a(k).e the king of Ur -
7: Ki-lul-la g!!-za-la Kilula guzala Kilula, the guzala-official,
8: dumu Ur-ba-?-ke4 dumu Urba? ak.e the son of Ur-ba? -
9: mu-na-dim kiib-ba mu.na.(n dim fashioned (this). The name
C
kiib.bi.a k) of this cylinder seal is:
10: lugill-gu l O Retug3 -nig-ag 5- lugal.gu .getug.nig. ag. "0 my king, let me keel? him
gg-ICa-ne ak.am.e alive at his ear of favor .
1 1 : gg-an-.ti-il ga.(L)n.til
12: mu-bi mu.bi
Commentary
3. Laga.ak.e, written Lagaki-ke4 ' The followers of the Falkenstein school would read
the first sign as Lagaa.
The .e at the end is the marker of the locative-terminative case. In all the previous
Lesson 22 25 1
instances of indirect objects and benefactives, the nominal phrase used the dative case,
marked in .ra. Occasionally, however, the locative-terminative case, marked in .e, is used
instead. This reflects a further spread of the usage of the locative-terminative; such a use of
the locative-terminative for the dative does not appear in earlier Sumerian.
The combination of the Ikl of the genitive marker with the lel of the locative
terminative is written with the ke4 -sign, just as is the combination of the Ikl of the genitive
marker with the I el of the ergative. (This ambiguity of the ke4 -sign can lead to confusion.)
4. Here and in line 1 1 , the root for "to live" is written !i-g, not tU. That is, the root is
written syllabically, and not logographically. The spelling nam-!i-g also occurs in the
Gudea inscriptions. Such spellings show us that the final 11/ of the root was indeed
pronounced in word-final position.
J
Lines 4ff. are a complicated genitive:
nam.til. Ulgi .ak.e
nitabkalaga
lugal. Urim.ak
9. Because the locative-terminative is used in line 3 in the sense ofthe dative, it is resumed
by na, the dimensional-prefix properly belonging to the dative. In its more usual uses, the
locative-terminative is either not resumed at all, or else it is resumed by ni - properly the
dimensional-prefix belonging to the locative.
There is no expressed direct object (patient); the votive cylinder-seal itself is the direct
object.
The last word in the line is an anticipatory genitive, beginning a new sentence: "of this
cylinder-seal, its name is ... ", that is, "the name of this cylinder-seal is ... ". This use of the
anticipatory genitive is similar to that seen in Text 15: alam-ba ... mu-bi-im, "the name of
this statue is ... ".
Line 9 includes the final word of one sentence, and the beginning word of the
following sentence. It is unusual in Sumerian orthography for one line to contain elements
of two different sentences. However, the space constraints in cylinder-seals occasionally
cause odd placement of signs within a line or case.
1 0- 1 1 . These lines express the actual name of the votive cylinder seal.
10. lugal-m l O is probably a vocative. As in Text 2 1 , the vocative is unmarked.
The use of nam to form abstracts has occurred several times. However, some
abstracts are formed with nig, instead of nam: ni g-Sag5 ' "goodness, favor". The original
meaning of nig is "thing", "something". It is relatively uncommon for nig to form such
abstract nouns. More commonly, it forms concrete nouns from verbal roots: g!!7 ' "to eat";
nig-g!!7 ' "food". Therefore, it might perhaps be best to regard the nig-Sag5 in Text 22 as
concrete, rather than as abstract. Sollberger, in fact, translates this expression (in another
text) as: "grace, favour (in a concrete sense, 'good things')" (1966: 1 58).
Unlike nam, nig is not used with nominal roots. That is, such forms as *nig-l!!gal do
not occur.
The transliteration ne assumes the interpretation as .ani.e, the possessive-suffix
followed by the locative-terminative case-marker, here with its more original meaning of
"at, by, through". A similar case occurred in Text 20, where the writing ne represented the
252 Lesson 22
possessive-suffix followed by the ergative case-marker .e. As discussed there, some
Sumero10gists would prefer to transliterate such phrases by ni: getug3 -nig-ag 5-g-ka-ni.
1 1 . g is the modal-prefix of the cohortative mood. This is used for positive wish for first
person, both singular and plural: "let me/us, may I1we". It is thus the first-person
equivalent of the third-person desiderative modal-prefix b.
The view of Sumerian grammar presented in this book assumes the presence here of a
conjugation-prefix 1. which has assimilated into the modal-prefix g. However, since a
writing such as *g-l seems never to occur, the assumption of such assimilation may be
simplistic. Therefore, the idea that a conjugation-prefix must be present in every finite
verbal form may need modifying.
One of the thornier problems in Sumerian morphology is the form of the root
appearing after the different modal-prefixes. In the singular, the cohortative g is always
used with the bamn!-root, whether the root is being used transitively or intransitively. (The
plural is unsure.) When the cohortative is used in the singular, there is no overt marking
for person; that is, the subject (or agent) is not marked. But the direct object (patient),
however, is expressed by a personal-affix immediately preceding the verbal root. Thus, the
.n before the verbal root here marks the patient, and hence the root ti-il must be interpreted
as transitive: "let me keep him alive/well".
The syntax of the personal-affixes of the cohortative (and of other moods) differs from
that of the indicative. In the indicative, the pre-verba1-root slot marks the agent in the
bamn!, but the patient in the mam. But in the cohortative, the agent is unmarked; the pre
verbal-root slot marks the patient, and the root is always in the bamty.
1 2. The enclitic copula is not used, although it was so used in the parallel expression in
Text 9: a1am-ba ... mu-bi-im.
Discussion: sign formation
In Lesson 9, the use of gunu-strokes to form new signs was discussed. A similar
device was the addition of eig-strokes to a sign. Whereas gunu-strokes consist of short
lines, eig-strokes look more like a kind of cross-hatching. For example, the da-sign is
originally a picture of a hand-upper-shou1der-arm: ,i=1. The word has such meanings as
"forearm, side", etc. To indicate .!i, meaning specifically "arm" (and then, by extension,
"strength"), eig-strokes appear on the part of the da-sign which approximately repre
sents the arm, yielding: .
The etymology of eig is unknown, but it must be connected with the word e, "bar
ley", whose cuneiform sign resembles these cross-hatchings.
- Syllabic writings
Syllabic writings of words usually written 10gographically (such as nam-ti-il for nam
til) are not common in the Ur III royal inscriptions; it is not sure why they occur. By way
of contrast, they are not uncommon in the two very large Gudea inscriptions.
Falkenstein believes that the relatively high frequency of such spellings in these two
Gudea texts shows that they were written down on the basis of a dictated text. That is, the
scribe(s) did not work from a written, already-prepared source, but rather the scribe(s)
Lesson 22 253
listened to the text being read, and wrote down the text as they went along. This idea does
seem like the best explanation for certain kinds of error which occur in the Gudea texts.
There is also some other evidence to support this view.
-Textual interpretation
The interpretation given above of the verb form in line 1 1 rests on the assumption that
the .n in the prefix chain refers to the patient. This view would probably be accepted by
most Sumerologists. However, other Sumerologists are less categorical in their thinking,
and would say that at our present state of knowledge, other possibilities cannot be
excluded. Therefore, this particular line has been translated as: "let me live by his ear of
favor", or even "let me make well his ear of favor". These translations reflect different
possibilities of understanding the .n: marker of the first-person intransitive verb, or marker
of the first-person transitive verb, or marker of the third-person patient, or reduced form of
the dimensional-prefix cross-referencing the locative.
- g and gae
The cohortative modal-prefix is /ga/, and the first person independent pronoun is
/gae/. Since one begins with /g/ and the other with /g/, it does not seem that they are
directly related to each other. On the other hand, the similarity in form, and the fact that
both are used for functions involving the first person, makes one pause. It is less easy,
however, to see such surface-level etymologies with the other modal-prefixes.
- Noun formation
The term zabar-dab5 0riginally meant something like "one who holds the bronze". The
word for "scribe", dub-sar, originally meant "one who writes a tablet". In Text 22, g!!-za
hi, "chair-bearer", occurs. A number of names of officials in Sumerian are composed of an
active participle and an incorporated direct object. Many were borrowed into Akkadian as
simple nouns: zabardabbu, tuparru, guzalU, etc.
- Seals
It is usually assumed that Kilula was the person who had this cylinder-seal fashioned.
He dedicated it to Meslamtaea, to bring life to Shulgi. Presumably Kilula gave it to Shulgi,
who then gave it to the temple. Thus, the cylinder-seal was designed to curry favor with
the king.
It is hard to say why the particular god Meslamtaea was invoked on this seal. Hallo
points out that "in private ex-votos inscribed on behalf of the king, it is not always certain
whether the deity involved is the personal god of the king or his donor" (1966: 1 37 n.53).
Without more precise knowledge of how such dealings took place, it is difficult to say
exactly what the name of the seal does mean. l!!gal, to illustrate part of the problem, can
mean "king, lord", referring to the god Meslamtaea. It can mean "king", referring to
Shulgi. It can even mean "owner", referring to the owner (which one?) of the cylinder-seal.
And, the locative-terminative (presumed ! ) at the end of the nominal phrase can have several
different values: "at, by, through", etc. A somewhat materialistic interpretation of the name
254 Lesson 22
would be: "let me make him well at his ear of favor", that is, "at his ear which hears and
grants favors". The sense is, "let him listen favorably to me", "make him accessible to me".
However, one can't help wondering whether Shulgi had an ear-ache.
- Votive seals
Votive-seals were basically non-functional, in the sense that they were not primarily
designed to be actually impressed upon written documents, as were the functional seals.
Rather, they were votive objects, in the form of a seal. There are very few examples of
votive seals actually being used (although it is always possible that this is an accident of
discovery).
Gelb says:
The main characteristic of the votive seals is that while they identified the
donor of the seal, they were not used by the donor but by the divinity to
whom they were offered. Certain seals can be used for purely ornamental
purposes but nothing would prevent the temple from employing them in
identifying and legal purposes (1977 : 1 12).
Votive seals are also generally larger than functional seals. And, whereas functional
seals are inscribed in reverse - so that the impression comes out correctly - votive seals are
not; they are meant to be looked at, not to be used.
Lesson 22 255
Text 22a
supplementary
256 Lesson 22
Sign-list and vocabulary
4. Ergativity
It is difficult to see all the manifestations of ergativity in the Ur III royal inscriptions,
because very few intransitive verb forms occur, especially in the maru. Here ergativity will
be summarized, even if some of the constructions do not occur in this corpus.
The following four sentences will illustrate: (1) the bamm-transitive; (2) the maru
transitive; (3) the bam!!-intransitive/passive; (4) the marfi-intransi tivel passive.
bam!!-transitive:
(1) The king built the house.
luga1.e e.0 mu.n.du.0
257
258 Lesson 23
manl-transitive:
(2) The king will build the house.
lugal.e e.0 Lb.du.e.0
bam!!-intransitive/passive:
(3) The king went.
lugal. 0 mu.gin.0
manl-intransitive/passive:
(4) The king will go.
luga1.0 Ldu.0
In both the bamm and the manl, in both the transitive and the intransitive/passive, the
nominal participants are marked in the same manner: the agent is marked in .e (the ergative
case-marker) and the patient is marked in .0 (the absolute case-marker).
Thus, looking at the manner in which the nominal participants in the sentence are
marked, Sumerian is an ergative language: the agent is marked in one manner (.e) , and the
patient in another (.0).
In terms of cross-referencing by the personal-affixes, however, the situation is
different. With a transitive verb, in the bamm the agent is cross-referenced by the personal
affix slot before the root. In the maru the agent is cross-referenced by the personal-affix
slot after the root. In the bam.!! the patient is cross-referenced by the personal-affix slot
after the root. In the maru the patient is cross-referenced by the personal-affix slot before
the root.
With an intransitive/passive verb, the bamm and the maru behave the same way: the
patient is cross-referenced by the personal-affix slot after the root.
This means that the cross-referencing system used in the maru behaves differently in
the transitive than it does in the intransitive/passive. Repeating the previous sentences in a
differen t order,
In the bamt.!! , the patients in (5) and (6) are cross-referenced in the same manner: by
the personal-affix .0 after the root. But in the maru, the patient is cross-referenced in two
Lesson 23 259
different ways: In (7), the patient is cross-referenced by the personal-affix slot before the
root. In (8), the patient is cross-referenced by the personal-affix slot after the root.
Thus, the patient in (7) (the direct object of the transitive verb) and the patient in (8)
(the subject of the intransitive/passive verb) are not treated the same. Therefore, the system
cannot be considered ergative, since the definition of ergativity is that these two patients be
treated the same.
This argument hinges on the definition of ergativity. In the bamn! and the maru, the
nominal participants are marked in an ergative manner. It is only in the system of cross
referencing that the two patients in (7) and (8) are treated differently from each other. If
one accepts the idea that ergativity does not just refer to the nominal markers, then we are
forced to say that the maru does not behave in an ergative manner. But since the bamtg
does behave in an ergative manner, Sumerian must be called a split ergative language, split
along an aspectual axis. However, if we look only at the nominal markers and not at the
system of cross-referencing, we can say that Sumerian is ergative, and not have to refine
this term (at least, in terms of an aspectual axis; there is some indication that Sumerian is
split along a nominal - pronominal axis).
(5) Personal-affixes
The interpretation of the personal-affixes presented here has been basically known
since Poebel. This interpretation was made much more explicit, and placed in an ergative
framework, by Michalowski (1982). While this interpretation seems to work for most
verbs, it is clear that it does not work for all of them. For example, a commonly occurring
sentence is: kiib PN.ak ib-ra, "the seal of PN was rolled". The verb apparently is intran
sitive/passive, as shown by the lack of an ergative case-marker .e, yet the .b in the pre-root
slot seems to cross-reference it. The proper analysis of such forms is still unclear to us; it
may be much more complicated than it first appears. More work remains to be done on
explaining and categorizing the exceptions which seem to occur with the personal-affixes.
(6) Root
Sumerian roots can be divided into two classes: nominal roots, such as lu, and verbal
roots, such as sar; there is no morphologically distinct class of adjectival roots. Adjectives
are to be regarded as participles (or something similar) of verbal roots. For example, gibil
can be used as an adjective meaning "new", from the verbal root meaning "to be new".
There is no canonical shape of the verbal root. Most of the verbal roots occurring in
the texts in this book are of the type CV (d.!!, g4), or of the type eve (g!!l2, pad). How
ever, there has been one eveve root (Wig), one of the type vev CM!!), one unsure (ba
ill) , and one borrowed from Akkadian (gi-in). Similarly, there is a fairly wide variation in
the shape of the nominal root. They have taken the form V @; CV Clli!.) ; VC (ur); eve
(gg); vev (utu); evev (dumu); veve (alam); eveve (temen); etc.
The root is unmarked for such categories as active - passive, transitive - intransitive,
causative - passive; etc. For example, til can mean "to live" or "to let live"; kur9 can mean
"to enter" or "to make enter", "to bring in". In Text 15, gub is used in the meaning "to
stand"; in Text 6, it is used in the sense of "to plant" a garden.
260 Lesson 23
B. Areas of disagreement
There are a number of disagreements and alternate explanations about certain features
of Sumerian grammar. Because a knowledge of these alternate views is presupposed by
Sumerologists, it is important to be at least familiar with the main differences from the
views presented here.
1 . Verbal phrase
The interpretation of the Sumerian verbal phrase presented in this book owes much to
the ideas of Gene Gragg (briefly sketched in [1968]). Two principal differing views are
those of Falkenstein and Jacobsen. Yoshikawa's views also differ considerably, but he has
not yet published a complete synthesis of his views.
(a) Falkenstein
His views are sketched in (1959), and particularly adumbrated for the Gudea texts in
(1978 2). His interpretation of the prefix chain is as follows:
Prllformative Konjugationsprllfixe Prafixe Verbalinfixe Wurzel
(1) (2) (3) (4) (5)
In general, Jacobsen's analysis tends to be much finer than that of other Sumerologists.
In some cases, vocalic (or consonantal) alternations which most Sumerologists would
regard as phonetic or morphophonemic alternation are regarded by Jacobsen as repre
senting different morphemes. He has been criticized for producing much too detailed a
segmentation of the morphology. On the other hand, Jacobsen's interpretations are based
on a close empathy for the texts; perhaps more so than any other living Sumerologist,
Jacobsen has a feeling for and understanding of the content of Sumerian texts, especially
literary texts.
(c) Yoshikawa
2. Maru-inflection
maru-forms such as l-sar-re have been analyzed here as Lsar.e.0. The root is sar,
which is a member of the affixation-class of maru-formation; .e is the maru-suffix; .0 is
the personal-affix cross-referencing the agent, that is, the third person marker.
Edzard believes that l-sar-re is to be analyzed as Lsar.e. sar is a member of the in
variable class of maru-formation, and .e is the third person marker.
The first analysis sees two morphemes after the root; the second sees one. It is not
easy to resolve this issue. The problem is partially the result of a lack of evidence. Far
fewer intransitive/passive verbal forms are preserved than transitive forms. Far fewer
present-future fornls are preserved than past. And, fewer first and second persons are
262 Lesson 23
preserved than third, and fewer plural forms than singulars.
Many of the examples preserved which show the greatest amount of grammatical
variation are attested in relatively late copies of literary texts. In these texts one must
always guard against misunderstandings by the Akkadian-speaking scribes who copied
down these texts. Similarly, the Akkadian-speaking scribes who drew up the grammatical
texts discussed in Appendix 2 sometimes include transitional or analogical forms, which
cannot be regarded as conforming to the standards of earlier Sumerian.
The problem is also caused by ambiguities in the writing system. Types of ambi
guities include: (1) The writing system cannot distinguish between a form such as i .sar.en
or L sar.e.en; they would both be written as l-sar-re-en. Isolated writings of the type i-sar
re--en do not resolve such ambiguities, because there are several possible explanations for
the full writing. (2) /d! is an amissable consonant. Therefore, a writing such as i-sar-re
can stand for i .sar.e, i . sar.ed, or L sar.e.ed. (3) The de-sign is ambiguous; it can stand for
de or for ne. This means that a writing such as i-sar-re--de can represent several different
possibilities: Lsar.ed.e., L sar.e.ed, Lsar.ene, L sar.e.ene, etc. It is ambiguities such as these
which make it diffic ult to correctly analyze the morphology of the maru.
3. Manl classes
The number of different types of formation of the maru from the bamt!! is unsure.
Yoshikawa's initial formulation showed three classes: (1) affixation (2) reduplication (3)
alternation. This scheme works well for the Ur In royal inscriptions, but this is partially
because only a very limited number of man1 forms occur in these texts. Yoshikawa himself
has said that his classification system will need expansion and revision.
Edzard modified and extended the system into five classes: (1) unchanging. Since
Edzard considers .e to be a third person subject marker, not a marker of the marG, this is
actually the same as Yoshikawa's affixation group. (2) reduplication (3) root-varying.
This is a sub-class of Yoshikawa's alternation class; the two roots are different, but phone
tically similar in some way. (4) alternation (5) irregular. These last do not seem to fit into
the first four classes.
As more progress is made in Sumerology, the last class will be further refined. Some
of the verbs which now seem irregular will eventually be shown to follow rules which are
not yet known. Whenever linguists study an unknown language, they are apt to see more
irregular forms than when they have been able to examine the data more thoroughly.
4. Normalform
There has been some discussion about the conjugation of the intransitive/passive verb
in Sumerian. Poebel believed that there was one conjugation for the bamt!! intransi
tive/passive, and a different conjugation for the marG intransitive/passive. Most modern
Sumerologists, however, believe that there is only one conjugation for both the bam n! and
the manl (that is, there is only one set of endings), although the particular root used (i.e.,
either bamn! or marG) will differ. This one conj ugation is usually called the "Normalform"
(even though this is a rather meaningless term). What follows is Poebel's reconstruction of
the bam n ! intransitive/passive and the marG intransitive/passive.
Lesson 23 263
For Classical Sumerian, Poebel reconstructs the intransitive/passive form of the bamt.!!
as:
That is, the man} intransitive/passive differs from the bamm intransitive/passive sole
ly in the presence of the element .ed. Poebel believed, however, that this sytem worked
only for Classical Sumerian. Under the influence of Akkadian, several analogical changes
took place in the system of the man} intransitive/passive, eventually yielding a completely
new paradigm:
The .ed element has disappeared; the third person singular is now marked by .e; and
the third person plural is now marked by .ene. Thus, the endings have been assimilated to
those of the maru transitive.
From the period between the time of the paradigm of the more classical period and the
time of the newer paradigm, several transitional forms are attested (for example, i-sar-re-de,
L sar.ed.e) for the maru intransitive.
PART THREE: APPENDICES
Appendix 1
History
265
266 Appendix 1
tablets represented numbers, but more likely they indicate items being counted; their precise
intepretation is unsure. Two from Tell Brak in northern Syria, found in 1984, in particular
are quite archaic looking. Based on archaeological criteria, these numerical tablets do not
seem to be any older than the pictographic tablets discussed above; rather, both numerical
and pictographic tablets occur in Uruk IV a.
A very archaic-looking tablet comes from Kish. Unfortunately, it is from an
uncontrolled context, and so it cannot be dated archaeologically. This tablet is frequently
referred to in popular literature about the Ancient Near East as being one of the very earliest
tablets known (if not the earliest tablet), but it has a fairly elaborate division into cases,
which makes it more likely that it is later than Uruk IV.
These early texts are undeciphered, and perhaps to some degree undecipherable.
Therefore, it is impossible to be certain about what language they are written in. There are
several reasons why these texts cannot yet be read:
- Even at this early date, the supposedly pictographic nature of the signs is not always
obvious. Most of the signs are already abstract; it cannot be determined what they were
originally meant to depict.
- Some of the abstract signs can be understood on the basis of knowledge of later
Sumerian. However, a fair number of the signs (perhaps 30%-50%) cannot be read or
understood. These are signs which eventually passed out of use, so that there is no later
grammatical tradition to provide information about their meaning.
- Most of the signs which can be understood are logographic; in theory, these can be
read in any language. A sign which is a picture of a mountain, for instance, could be read
as "mountain", "Berg", adfi, kur, etc.
- There do not appear to be any syllabic signs. This means that no grammatical features
can be seen; for example, there are not any case-markings on nouns. There do not appear
to be any verbs at all. Writing at this stage was a highly mnemonic device.
- These are mostly administrative records, sometimes very short - occasionally, just a
few signs long. It is very difficult to understand such texts out of context, that is, without
knowledge of the administrative framework which produced these texts.
In spite of these problems, most scholars think that these texts are written in Sumerian.
The main reason is because texts have been preserved from the later Uruk III stratum,
which is known to be Sumerian. Since archaeologists see a cultural continuum between
Uruk IV and Uruk Ill, it is reasonable to assume that the same language is present in both
strata. In addition, Powell has argued that the system of metrology used in these early texts
seems to be the same system used in clearly Sumerian texts:
The system of numeration deducible from the notation present on Uruk
IVa/III tablets makes it virtually certain that these tablets are written in
Sumerian and, ipso facto, highly probable that the inventor of the pictorial
writing system was also a Sumerian (1981 :423).
These tablets are being studied by Margaret Green and Hans Nissen (a student of
Falkenstein), and are in the course of publication (preliminary discussions Nissen 1985,
1986; first major publication Green and Nissen 1987). Although most earlier scholars
differentiate between Uruk IVa and Uruk Ill, Nissen subsumes both into one category,
History 267
"Archaic Texts". (The figure of "4000 texts" cited above therefore includes tablets from
both Uruk IVa and Uruk lll.) Nissen estimates that about 85% of the texts are economic
records, and about 15% are lexical lists. He is more optimistic than most scholars about the
possible decipherment of these texts, believing that he can identify about 700 of 1000
different signs, and that the texts are "possibly" written in Sumerian. He bases his
arguments partially on the continuity between the early lexical lists (unknown at the time of
Falkenstein's publication) and later, well-understood lexical texts.
Nissen thinks that the texts from Uruk can be divided roughly into two classes, one
representing an "early stage of the script", and one a "younger stage of the script" (these
two divisions do not exactly correspond to the traditional Uruk IVa/Uruk III divisions).
The latter, consisting of most of the tablets found since Falkenstein's publication, are more
amenable to analysis.
The fact that many of the signs in these early tablets are already abstract has led many
scholars to assume that there was some previous development behind the signs. That is,
these tablets do not represent mankind's first attempt at writing. Several different
hypotheses have been proposed:
- The Sumerians may have borrowed their writing system from some other people,
perhaps some distance away from Mesopotamia. This is not impossible. It has often been
argued that the Sumerian writing system does not fit the Sumerian phonological system
very well; this might imply that the writing system was created for a different language.
This particular theory has been around for many years; it is obviously very difficult to
prove.
- Earlier writing may have been on perishable material, such as wood, or animal skins,
or palm leaves, etc. There are parallels to such practice from later Mesopotamia, and from
Arabia around the time of Muhammad. This is also a rather old theory, but it is also
virtually impossible to prove.
- In a series of articles beginning in 1977, Denise Schmandt-Besserat has argued that
the earliest "precursor" of writing was clay "tokens", which have been found at various
sites throughout the Ancient Near East, starting from the early Neolithic. Writing
originated in a conceptual leap, from the use of physical tokens, to the use of symbols to
represent these tokens: "The substitution of signs for tokens was no less than the invention
of writing" ( 1986:37). She envisages the following stages (1986:35):
(1) 8000 BC appearance oftokens
(2) 3250 BC clay envelopes hold tokens of particular transactions
(3) 3200 BC signs are impressed on the surface of envelopes
(4) 3 100 BC clay tablets appear with impressed and incised signs
Powell agrees:
Cuneiform was invented in a short period of time around 3000BC by a
citizen of the Sumerian city of Uruk. ... It arises conceptually out of the token
system described by D. Schmandt-Besserat. ... The pictorial ancestor to
cuneiform writing was invented as a conceptual whole during the time period
represented by the Uruk IV-Ill archaeological strata (198 1 :419-420).
268 Appendix 1
earlier. Other texts turned out to be previously unknown compositions. For example, one
is a collection of temple hymns. Most of these texts are scarcely intelligible; not much is
known about literary Sumerian of this period.
Thus, the primary importance of Tell Abu Salabikh lies in the existence of literary texts
from the middle of the third millennium B C. Since these discoveries, scholars have
recognized fragments of literary texts among some tablets which have been known for
many years. For instance, some of the Fara texts are fragments of proverbs which are
known from later proverb collections.
The texts from Tell Abu Salabikh are also important, because a number of the literary
texts (and lexical texts) have colophons of the sort: "so-and-so wrote". It is not known
what the word "wrote" means here exactly: Does it mean that the scribe "composed" the
composition, or that the scribe "copied" the text from a master tablet, etc. However, what is
interesting is that a number of these scribes have demonstrably Semitic personal names.
It is difficult to date the intrusion of Semitic-speaking peoples into Mesopotamia, on
linguistic or other grounds. The first evidence is usually thought to be the presence of
Akkadian loan words in early Sumerian. These loan words are difficult to evaluate, how
ever, because it is not always certain which way the borrowing went, or whether a third
language may have mediated a word, etc.
The Semitic names in these colophons are thought to be the first real evidence of
Semitic-speakers in Mesopotamia. If the Fara texts and the Tell Abu S alabikh texts are
dated to about 2600 BC, that gives a terminus ante quem for the arrival of Semitic speakers,
but it does not say anything about how long they might have been present in Mesopotamia.
If they had already become scribes, they must have been there for some time, since they had
worked themselves into the intellectual life of the community.
Most of the texts of this stage come from Lagash, from a period known as the "First
Dynasty of Lagash". Besides the usual administrative, economic, and legal texts, there are
a fair number of royal and private inscriptions. There are also some letters, and even a few
literary fragments are now known. Royal inscriptions are also known from other sites.
The end of this period corresponds to the rise to power of the Semitic-speaking
Dynasty of Akkad (2334-2 1 54 BC). As mentioned above, Semitic-speaking peoples must
have been present in Mesopotamia for centuries before the time of Sargon, the founder of
the dynasty (ruled 2334-227 8 BC). It must be presumed that Mesopotamia was bilingual
during this time, at least to some degree. However, with a Semitic-speaking dynasty in
power, Sumerian gradually started to move into second place.
A recent addition to the corpus of texts known from this period are the texts from Ebla
(in northern S yria, therefore from outside of the Sumerian-speaking heartland). To date,
upwards of ten thousand texts and fragments have been discovered. The texts are in both
S umerian and Eblaite; until the material is better studied, it is not sure which language
predominates. Early accounts said that perhaps ninety percent of the texts were written in
Sumerian; this is probably much too high a figure. The problem is that the texts written in
Eblaite are couched in a Sumerian orthography, utilizing a large number of Sumerian
270 Appendix 1
logograms.
Most of the texts found at Ebla are administrative or economic, chiefly concerned with
the metal and textile industries. However, there are also lexical lists; some are Sumerian
lists known from later periods, others are bilingual Eblaite-Sumerian texts. There are a few
literary texts in Eblaite (mostly incantations); these are extremely difficult to understand.
The existence of possible literary fragments in Sumerian is disputed.
It is still too early to assess the Sumerian texts from Ebla. It is clear, however, that
much new information is present. For example, the bilingual lexical texts include Sumerian
words and expressions not elsewhere attested.
Although Sumerian was on the defensive in the face of Akkadian, it enjoyed a strong
- albeit brief - revival under the kings of the Vr ill Dynasty (2 1 1 2-2004 BC). This is the
period from which the most tablets of all have been preserved. There are texts from many
sites, including Vr itself, Drehem, Lagash, Larsa, Nippur, and Vmma. There are literally
thousands and thousands of mostly economic documents, as well as inscriptions, letters,
and other types of texts. Also, more and more tablets with literary texts are being dated to
this period.
From some time before the reign of Vr-Nammu (the founder of the Dynasty), there are
a fair number of inscriptions from the reign of Gudea, the local ruler of Lagash. Many of
these are inscribed on statues of Gudea himself. There are also two large cylinders of his,
inscribed with a very long building hymn. The largest ("Cylinder A") is almost one
thousand lines long; it is apparently the longest connected Sumerian inscription.
The dynasty of Gudea is referred to as "Lagash II". The chronology of Lagash II is
unsure; some see it as roughly contemporaneous with Vr Ill, but most view it as following
immediately upon the Old Akkadian period. In any case, the language of the Gudea texts is
more or less the same as that of the Vr III texts.
It is not known when Sumerian ceased to be a spoken language; this is a current topic
of discussion among Sumerologists and Assyriologists. Vsually assumed to be spoken
during the Vr III period, it was under the greatly increasing influence of Akkadian. Some
scholars use the figure 2000 BC, others 1 900 BC, for the date when Sumerian ceased to be
spoken, but this figure is rather arbitrary. The language continued to be spoken by ever
smaller groups of speakers, and it is impossible to say when the last speaker of Sumerian
died. What is usually meant by the question "When did S umerian die out?" is "When did
the native language of the people who produced the texts we have cease to be Sumerian?"
Pockets of native speakers of Sumerian may have continued for some time, but without
producing any texts.
Other scholars have argued for an earlier death. Jerrold Cooper has said that
"S umerian as a spoken language was in all probability dead or nearly so in Ur Ill"
( 1 97 3 : 24 1 ). His argument is based on the types of documents preserved during the Ur III
period. Both Kienast ( 1 98 1 a) and Michalowski ( 1987) essentially agree.
Jacobsen, on the other hand, says "We therefore assume that Sumerian was still
spoken as everyday language in the south in the Ur ill period and a major part of the Isin-
History 271
Larsa period as well ( 1988: 124). Liebennan has stated that there is some evidence to show
that "Sumerian was spoken during the Old-Babylonian period" (1979:27).
V. Post-Sumerian (2000 BC-lOO AD)
This is occasionally divided into the following subdivisions:
Early Old Babylonian 2000- 1800 BC
Later Old Babylonian 1 800- 1600 BC
Post Old Babylonian 1600-
After the Early Old Babylonian period, Sumerian was essentially dead as a living
language. However, it continued to be taught in the schools as a language of culture, and as
a language of religious importance. The parallel has frequently been made with the role of
Latin in the Roman Catholic Church: Latin is still written, and even to some degree
spoken; hence, it is "living", even though it is not spoken as a native language of anyone.
Recently, Vanstiphout has used the tenn "Standard Sumerian" to mean
the language used in the literary documents of the Ur III and Old Babylonian
periods. ... This language is a literary and therefore written fonn, taught in
school for educational and literary purposes ( 1985: 1 ).
The majority of Sumerian literary tablets which have been preserved are from the Early
Old Babylonian period. However, there are also original texts written in Sumerian from
this period; examples include royal inscriptions (alongside those written in Akkadian), and
hymns written in honor of some of the Old Babylonian rulers.
Although the original composition of most Sumerian literary texts was in Sumerian,
by native speakers of Sumerian, the native language of the scribes who copied down the
literary texts during this period was Akkadian, not Sumerian. This led to a strong linguistic
influence of Akkadian upon Sumerian, to the extent that the literary texts contain features
which would appear to be "wrong" by the rules of Classical Sumerian grammar.
Michalowski, for example, has spoken of the "profound changes in grammar evident in the
Old Babylonian literary texts" (1980a:9 1); "during the Old Babylonian period
Mesopotamian scribes wrote Sumerian utilizing a profoundly different grammar, much
influenced by Akkadian" (1980a:86 n.3). The extent of deviation from the nonn varies
from one particular text to another. Inanna's Descent, for example, is pretty good Su
merian, with only a few "wrong" verbal fonns. In Gilgamesh and Agg, on the other hand,
there are more "wrong" verbal fonns than "right" ones.
It is, of course, always possible that fonns which we regard as "wrong" are in fact
"right", but our understanding of Sumerian grammar is not yet sophisticated enough to
correctly interpret such fonns. J acobsen has emphasized this methodological point:
Once it has been decided that our sources are generally suspect it becomes
natural to see all unexpected and difficult features as due to corruption,
without seriously considering the possibility that our own limited and rough
knowledge might be at fault and need nwision . ... The essential thing is to be
slow to dismiss difficulties with the easy assumption of mistakes by the
Ancients ( 1988: 1 25-1 26).
272 Appendix 1
Sumerian continued to be written right down to the Christian era. These very late texts
are either cultic or astronomical. There are even a few Sumerian texts (including portions
of canonical lexical lists) written in Greek characters. The very latest cuneiform texts
preserved are several astronomical almanacs, written in Akkadian with a great number of
Sumerian logograms for technical terms. The latest of these can be dated by internal criteria
to the year 385 of the Seleucid Era, corresponding to 74175 AD.
As mentioned above, this periodization (and most others) is to some degree based on
external (historical and political) criteria, not on purely linguistic criteria. Jacobsen has
proposed a different scheme, based on linguistic criteria (without yet assigning precise
dates), while emphasizing the fact that the paucity of the data prevents overly-fine
subdivisions:
I Archaic
IT OldSumerian
ITI Standard Sumerian (beginning with Naril.m-Sin of Akkad)
IV LateSumerian
Appendix 2
Mesopotamian Sources
Much of our knowledge of Sumerian derives from the intellectual activity of the
Mesopotamian scribes themselves. This section describes some of these Mesopotamian
sources.
Lexical lists
It is especially in the area of lexicography that modem Sumerological studies depend
on native sources. From a very early period, the Sumerians began to compile "lexical lists".
These early texts were monolingual, consisting simply of lists, usually of words for
semantically related things: lists of names of fishes, of professions, of stones, etc.
Although most lexical lists are loosely arranged according to subject, others are organized
according to graphic shape, or even according to phonological shape. Fragments of such
lists occur among the very earliest Sumerian texts which have been preserved.
These early texts were the product of Sumerian scholars, originating in the Sumerian
scribal school system. Lexical lists become more and more common, however, beginning
about with the Old Babylonian period. At that time, Sumerian was in the process of
completely dying out as a spoken language - if it had not already done so. By the end of
the Old Babylonian period, if not earlier, Sumerian was only spoken in the schools. These
later lexical lists are a product of the Mesopotamian scribal schools; their purpose was to
aid the Akkadian-speaking scribes in their study of Sumerian.
By the late Old Babylonian period, many lexical lists assumed what is often called
"canonical" status; that is, they became standardized in content and in form. There are
about a dozen such "canonical series". Some are monolingual in Sumerian, like the earlier
texts, but most are bilingual; they have a Sumerian word in the left-hand column, and an
Akkadian equivalent in the right-hand column. Some have three columns: a phonetic
spelling of the sign; the S umerian logogram; and the Akkadian meaning.
Many of these series are quite extensive. One of the largest and best-preserved is
known (both to us and to the ancient Mesopotamian scribes) as "urS -ra = bubullu", after its
first entry. In its canonical form, this series occupied 24 large tablets, totalling about
1 0,000 entries. Civil has called it an "inventory of material culture" ( 1 976: 1 25). He
describes its contents as:
273
274 Appendix 2
The first entry of this series has Ufs-ra (the Sumerian word for "interest-bearing loan")
in the left-hand column, and the Akkadian gubullu (with the same meaning) in the right
hand column.
Lexical lists such as these help us to detennine the meaning of Sumerian words. Some
of the lexical lists go even further, and enable us to determine the reading (that is, the
approximate phonetic rendering) of a certain sign. For example, there is a relatively late
lexical series known as "diri", which in its canonical fonn occupied seven tablets, with
more than 2,000 entries. This series was used to give the pronunciation of compound
logograms, that is, logograms whose reading cannot be inferred from the individual parts
(such as zabar, written with the UD-KA-BAR signs; without lexical lists, it would be vir
tually impossible to deduce that the pronunciation of these three signs was Izabar/). For
this reason, such compound logograms are often referred to as "diri-compounds".
In diri, the pronunciation of the logogram under discussion is given in the far left-hand
column, using a restricted number of syllabic signs. Then comes the logogram in question.
Then comes the name of the sign (at least as early as the Old Babylonian period, the
Akkadian scribes gave names to the individual signs). Finally, the last column gives the
meaning of the sign, in Akkadian. A typical entry reads:
di-ri diri si-y-ku wa-at-ru
This tells us that the sign is read Idi-ri/. Graphically, this sign "looks like" the si
sign ( ) followed by the -sign ( Tf ). (At least, in this period of cuneifonn writing.
Originally, the diri-sign may have had no connection at all with either the si-sign or the
sign. However, by the Old Babylonian period, when signs were becoming more linear, it
happened to assume a shape looking like the si-sign followed by the -sign.) Because of
this external similarity, the Akkadians named this sign "the si of ", that is, si.a. (k). Finally,
the last column gives the Akkadian translation, "excess" or "extra".
Copies of these canonical texts have been found all over the Near East, not just in
Mesopotamia. There are also somewhat similar texts, but not of any canonical status, both
from Mesopotamia and from outside of Mesopotamia. Their function was the same, to aid
local scribes in their mastery of S umerian (or of some other language).
Some of these non-canonical texts are bilingual, some are trilingual, and some even
quadrilingual. For example, from Boghazkoy in Asia Minor there are several Sumerian
Akkadian-Hittite trilingual vocabularies. The native language in Boghazkoy was Hittite;
these texts were designed to help Hittite scribes in learning both Sumerian and Akkadian.
From Ugarit, there is a quadrilingual "vocabulary". It has entries in Sumerian, Akkadian,
Sources 275
The tablet is equating the Sumerian comitative case-marker da with the Akkadian
preposition qadu, meaning "with".
Lines 405-408 of the same tablet read:
The Sumerian cohortative modal-prefix gg!, and three morphophonemic alternants of the
desiderative modal-prefixb, are all "translated" as the Akkadian desiderative-marker lu.
The NBGT texts occasionally add scribal comments or annotations, in both Sumerian
and Akkadian. Examples include the Sumerian word AN-TA, "prefix", and the Akkadian
expression a iten, "singular". These kinds of annotations do not appear in OBGT.
OBGT and NBGT are rather extensive. One might think that these texts could furnish
a key to Sumerian morphology. Unfortunately, it is not so. These texts are all relatively
late. They represent Akkadian-speakers' understanding of Sumerian. However, these
Akkadian-speaking scribes did not always understand Sumerian grammatical categories
and distinctions. For example, in the passages from OBGT cited above, a difference in
Sumerian conjugation-prefix was equated with a distinction in Akkadian stem. It is
difficult to say how accurate an equation this is. Or, Black has pointed out that OBGT V
makes a consistent distinction between the first person suffix lenl, written -en, and the
second person suffix lenl, written --en (1984:7). Black thinks that this might indicate a
difference in pronunciation, but more likely it is a purely graphic distinction.
Regarding the Sumerian and Akkadian grammatical terms which sometimes occur as
annotations, Black has also said that there is
a growing body of evidence that the scribes responsible for introducing the
grammatical terms into the grammatical analysis texts sometimes misun
derstood their meaning (or misunderstood the texts into which they were
introducing them) (1984:90).
There is also a certain amount of systematization and schematization in these tablets.
But at the same time, there is no unified method of organizing the data. As mentioned
above, OBGT has over 200 lines of gm:, but it is not always easy to follow the principles by
which these forms are organized. Other OBG T texts follow their own organization.
Like the lexical lists, there also occur mechanical errors due to problems of text
transmission. There are also scribal conventions and abbreviations. And as mentioned
earlier, there was undoubtedly a large oral component which accompanied the study of
these texts, a component which is no longer accessible. These problems (and others) mean
that although these grammatical texts are a font of useful information, this information
cannot be used uncritically. These texts cannot be viewed as an exact reflection of
Sumerian of the Ur III or earlier periods. Black has said that "in some cases it seems that
we know Sumerian better than the compilers, or copyists, of our texts" (1984:7). This is
especially true regarding the Sumerian aspectual differences, and also regarding Sumerian
278 Appendix 2
causative sentences, a type of sentence heavily dealt with in OBGT. (Similarly, Black
believes that certain Akkadian forms seen in OBGT were "especially concocted to set
against certain Sumerian forms" [ 1984:29].)
Thus, although Jacobsen is undoubtedly correct in saying that the OBGT
constitute without question the most important single group of sources both
for the history of grammatical studies generally and for our understanding of
Sumerian grammar specifically so far known ( 1 956: 1 *),
these texts must be approached with caution. Jacobsen adds that "the immensity of the
number of problems raised, and the relative insufficiency of our present knowledge of
Sumerian becomes only too clear as one approaches the texts in earnest" ( 1 956:2*).
As the above quotes exemplify, scholars vary in their estimation of the worth of these
texts in reconstructing Sumerian grammar. Jacobsen, for example, sees them as extremely
important; Krecher, on the other hand, in his study of the conjugation-prefixes containing
an fmf element, found them to be of little value ( 1985:34).
Finally, Civil has pointed out:
To my knowledge, the fundamental question: are the grammatical texts des
criptive or prescriptive ? has never been formulated ... at least in print, al
though there is widespread skepticism about their descriptive adequacy
( 1986:72).
For a long time, it was believed that OBGT texts represented mankind's first attempt to
formulate "paradigms"; these texts antedate the grammatical studies of the Indians by over a
millennium. However, what may well represent the first attempt known to organize verbal
forms into a paradigm is now found among the Ebla tablets. From a rather variegated
monolingual lexical text (TM.75.G.2260), the following lines appear:
(line 1 2) : in-na-sum
I-na-sum
nU-l-na-sum
hi-na-sum
ba-til
nu-til
in-til
hi-til
For several reasons, one might be hesitant about considering these lines to be a
"paradigm". But they do seem to indicate that the "mind" of the scribe(s) was heading in
that direction. And this tablet antedates the OBGT texts by some 500 years.
The function of the lexical texts (and the smattering of what might be called a
"grammatical text" just discussed) at Ebla was similar to the function of the lexical and
grammatical texts found throughout Mesopotamia and the Ancient Near East in general: to
enable scribes to master Sumerian. For the Eblaite scribes, however, Sumerian was a
language still being spoken. For the Akkadian scribes of the Old Babylonian period,
Sumerian had virtually ceased to exist as a spoken language, and was only a language of
the schools.
Sources 279
OBGT and NBGT were published in MSL IV (1956). These texts were prefaced by
Jacobsen with a discussion of the Sumerian verbal system as reflected in these texts.
Jacobsen has also written a very interesting article intended for a more general linguistic
audience, discussing the system of paradigms seen in OBGT and NBGT (1974). Black
(1984) has written a book especially on these grammatical texts, and on the philosophy of
language which they represent; this work also has much incidental discussion of various
aspects of Sumerian morphology.
Very recently, Civil et al have published some "Middle Babylonian Grammatical
Texts" ( 1986). These have not yet been fully studied.
Syllabic S umerian
In addition to the "standard" or "normal" orthography and spelling of Sumerian, there
is a certain amount of what is called "syllabic Sumerian" or "phonetic Sumerian". Standard
Sumerian is written using a combination of logographic and syllabic signs. Syllabic
Sumerian, however, is written using only syllabic signs. For example, the standard Su
merian orthography for a locative phrase, kalam-ma, is written in syllabic Sumerian as ka
la-ma.
There are not a great deal of texts in syllabic Sumerian; they are all relatively late.
Interestingly, not many syllabic texts come from the Mesopotamian heartland; they are
mostly from northern Babylonia, or farther afield. The practice probably originated in the
scribal school system as a device for the scribes to cope with the difficulties of standard
Sumerian orthography'. In the case of syllabically-written incantations and liturgical texts,
the purpose was probably to aid in correct recitation.
Unlike the lexical and grammatical texts discussed above, there is no standard or
canonical system of writing syllabic Sumerian; it varies to some degree from text to text.
Much of it was probably produced on an ad-hoc basis, to deal with particular texts.
Since syllabic Sumerian is an attempt to reproduce spoken Sumerian, it should reveal
some of the features not shown in the normal orthography. Thus, one might think of it as
another key to unlocking Sumerian morphology. Unfortunately, it is very difficult to
understand syllabic Sumerian, even more difficult than it is to understand Sumerian in
standard orthography. The reason is precisely because standard Sumerian masks certain
phonetic problems, such as morphophonemic alternation, contraction, assimilation, etc.
When such phenomena actually show up in syllabic Sumerian, it is often difficult to
untangle the forms. Even in cases where the same text is preserved once in standard
orthography and once in syllabic orthography, the phonetic relationship between the two is
not always easy to see.
A relatively simple case is the writing at-ta, for standard an-ta, "from the sky", or
"from above". Should it be assumed that the standard Sumerian was also pronounced
latta/, and that the written form an-ta is a morphographemic or historical writing? If so,
should the an-sign be transliterated by an at-value? Perhaps in early Sumerian, the word
was indeed pronounced lanta/, but an assimilation took place in later Sumerian, producing
latta/. How can this change be dated? On the other hand, perhaps such a writing as at-ta
reflects the Akkadian assimilation of nasals, and doesn't say anything about Sumerian.
280 Appendix 2
Even in this one simple instance, one can think of several variables which must be
taken into account. But consider an even more complicated case. From the root hlL
meaning "to be confused", there appears a man1 form in standard orthography as ba-bir
bir-re (this shows formation of the marii by means of both reduplication and the marii-suf
fix .e). This appears in syllabic orthography as ba-bi-ib- presumably representing
Ibabibre/. Does this mean that the writing in standard orthography, ba-bir-bir-re, should
also be understood as representing Ibabibre/, and that this is a morphographemic or
historical spelling? How should it be transliterated?
Such examples illustrate the extent to which phonetic processes are masked by
standard orthography, and they show the difficulty in interpreting the syllabic forms. And
since syllabic Sumerian varies to some degree from text to text, it is difficult to generalize
about what is seen. All syllabic texts are rather late, from the Old B abylonian period or
after. They thus reflect a stage when Sumerian was no longer a spoken language, so to
some degree the phonetic differences that appear may be conditioned by the Akkadian
language of the scribes.
Very recently, however, there have been found at Ebla syllabically-written versions of
lexical lists. Civil has called their existence "a most unexpected surprise which opens a
new chapter in the understanding of the earliest lexical compilations and provides
phonological data for the oldest stages of Sumerian" ( 1 982: 1 ) . These syllabically-written
lexical texts are very difficult to interpret, and only preliminary work has been accom
plished.
Much more work, in general, remains to be done on syllabic Sumerian. The two
examples given above show the kinds of information which such texts can provide. A
more thorough investigation might help in solving some of the perplexing problems
encountered in the Sumerian writing system.
The Emesal dialect of Sumerian is written in a mixture of standard (i.e. , mostly
logographic) Sumerian and of syllabic Sumerian. For example, the word for "lady" in
Emesal is gaan, corresponding to Main Dialect nin. Sometimes Emesal texts simply use
the same nin-sign; it is assumed that the "reader" will know enough to render the nin-sign
as the Emesal equivalent gaan. More frequently, however (at least, with this particular
word), the word is spelled syllabically, gf!-a-an.
An Emesal vocabulary has also been preserved, some 1 77 lines long. It gives the
Emesal form in the far left-hand column; the Main Dialect form; and an Akkadian
translation in the far-right column. A typical example is line 96:
This tells us that the adjective "wide", Akkadian rapu, which is written with the dagal-sign
in Main Dialect, is I damall in Emesal.
Bilinguals
A certain number of "bilingual" texts have come down to us. These are of two main
types. In "interlinear" texts, a line of Sumerian is followed by a line of Akkadian. There
are many incantations of this type. Other texts are written in "parallel columns", with the
Sources 28 1
There is no standard sign-list for Sumerian. However, the sign-lists of Borger and
Labat, even though based on Akkadian values, are still useful for Sumerian. Labat is
especially helpful for the study of the palaeography of the signs, that is, the variation in
their basic shape throughout time. Borger provides some information about Sumerian
grammar, and also a certain amount of bibliographic material. The two volumes of
Ellermeier which have appeared to date are useful in sorting out inconsistencies in
published transliterations.
At the moment, there is no up-to-date dictionary of Sumerian. For a number of years,
the University of Pennsylvania has been preparing just such a project, the Pennsylvania
S umerian Dictionary (PSD; cited in the Bibliography under Sjoberg [ 1 984]). As of this
writing, the only volume which has appeared so far is for the letter "B". The glossary of
Delitzsch, although dating to 1 9 14, is still one of the most useful single-volume dic
tionaries, although not the easiest to use. There are a few glossaries to specialized bodies of
texts. One of the more recent is Behrens and Steible ( 1983). Unfortunately, even semi
serious lexicographical work usually means looking at the glossaries and indices of many
different text editions. For the beginning student, perhaps the most useful of these is
Sollberger ( 1 966). Professional Sumerologists keep very large and detailed files on
S umerian words. The core of the PSD, for example, is Ake Sjoberg's collection of over
500,000 dictionary entries, which he started in 1 949.
a (Text 1 )
arm, strength (Text 22)
a ... fU to dedicate a votive object (Text 4)
ab (Text 17)
abzu apsu (Text 14)
ad (Text 1 9)
ad-da father (Text 1 9)
Adab Adab (GN) (Text 20)
g (Text 8)
i!gi!2 (Text 8)
alam statue (Text 1 5)
am (Text 1 1 )
ama mother (Text 1 5)
amar young bull (Text 1 3)
d
Amar- Zuen Amar-Sin (PN) (Text 1 3)
an (Text 1 8)
An An (DN) (Text 6)
an heaven (Text 1 )
283
284 Appendix 3
id (Text 1 9)
id (h) river, canal (Text 5)
il (Text 22)
im (Text 1 5)
in (Text 7)
Inanna Inanna (DN) (Text 2)
iq (Text 1 9)
ir (Text 1 4)
is (Text 22a)
IS-ku-un-dZuen Ishkun-Sin (GN) (Text 22a)
isib (kind of priest) (Text 1 9)
ka (Text 5)
kalag (kala) to be mighty (Text 2)
kalam land (Text 17)
kar quay, pier; market place (Text 17)
Kar-zid-da Karzida (GN) (Text 17)
ke4 (Text 1 ; Discussion, Lesson 1 )
ki (determinative following GNs) (Text 1 )
ki place, eanh (Text 6; Text 1 2a; Text 1 4)
g
ki ... 2 to love (Text 8)
Ki-en-gi Sumer (GN) (Text 2)
Ki-lul-la Kilula (PN) (Text 22)
Ki-uri Akkad (GN) (Text 2)
kiri6 garden (Text 6)
kisib cylinder-seal (Text 22)
ku (Text 22a)
kug Gill) bright, pure, holy (Text 1 7 ; Discussion, Lesson 1 8)
kur mountain; highland; foreign land (Text 5)
kur to change (Text 1 5)
kurs Ck!! s ) (Text 1 2)
kur9 (ku4 ) to enter (Text 17)
la (Text 4)
la (Text 7)
la to hold, to lift, to carry (Text 22)
Lagas Lagash (GN) (Text 22)
lal honey (Text 1 6)
lam (Text 22)
Lamar DN (Text 1 1 )
le (Text 1 3)
li (Text 1 1 )
li (Text 1 7)
lil air, wind (Text 5)
limmu2 four (Text 1 0)
Glossary 287
nu not (Text 1 6)
numun seed; offspring, progeny (Text 1 5)
nun (Text 5)
nun prince, noble (Text 1 6)
pad (w) to find, call, reveal (Text 1 3)
nill: (Text 1 7)
ra (Text 5)
re (Text 6)
re6 (Text 1 6a)
ri (Text 1 9)
ru to send (Commentary, Lesson 4)
sag head (Text 9)
sag slave (Text 2 1 c)
Sag- dNanna-zu Sagnannazu (PN) (Text 2 1 c)
sag-lis supporter, sustainer, patron (Text 1 3)
sagi! (kind of priest) (Text 2 1 c)
sar (Text 1 2)
sar to write (Text 1 9a)
sig9 (si) to be narrow (Text 1 5)
sikil to be pure, clean (Text 6)
sipad (sipa) shepherd (Text 1 9)
siskur2 sacrifice (Text 1 6)
su (Text 1 1 )
sud4 13) to be long (Text 1 7)
sag4 (sa.) heart (Text 1 8)
sags (g6 ) to be good (Text 22)
S ara
2 Shara (DN) (Text 1 9)
se (Text 4)
silig to cease (Text 1 6)
su (Text 1 8)
su hand (Text 1 9)
su ... ur to erase (Text 1 2)
d
S u- Zuen Shu-Sin (PN) (Text 1 8)
suI (Text 1 0)
S ul-gi Shulgi (PN) (Text 1 0)
ta (Text 1 7)
temen (Text 9)
ti (Text 1 9)
til to put an end to (Text 1 5)
til Cti) to live (Text 4; Commentary, Text 22)
tu (Text 1 9)
turn (Text 1 8)
tur (Text 1 6a)
Glossary 289
!! (Text 1 1 )
!! and (Text 1 6)
ub (Text 1 0)
ud ) day (Text 1 2)
ul remote, distant (Text 1 7)
un (Text 17)
Unug (Unu) Uruk (ON) (Text 9)
ur man, warrior (Text 1 )
Ur-ba- ? Urba? (PN) (Text 22)
d
Ur- Lamar Ur-Lamar (PN) (Text 1 9a)
d
Ur_ Nammu Ur-Nammu (PN) (Text 1 )
Ur-dNanibgal Ur-Nanibgal (PN) (Text 2 1 b)
Ur-dNin-grr-su Ur-Ningirsu (PN) (Text 1 1 )
ur (Text 1 2)
Uri (Text 2)
Urims (Uri s ) Ur (ON) (Text 1 )
us (Text 1 3)
utu sun (Text 17)
Utu Utu (DN) (Text 4a)
zabar bronze (Text 1 1 ; Discussion, Lesson 20)
zabar-dab S (kind of official) (Text 1 1 )
zid (zi) effective, true (Text 1 7)
zu (Text 1 3)
zu your (Text 2 1 b)
zu to know (Text 2 1 c)
Zuen Zuen (DN) (Text 1 3 )
"5" (Text 2 1 a)
Appendix 4
Bibliography
Abbreviations
Works Cited
van Aalderen, C.T. 1 982. "Some observations on ergativity and Sumerian". Orientalia
Lovaniensia Periodia 1 3 : 25-44.
29 1
292 Appendix 4
Halton (ed), Langua,ge, Literature, and History: Philological and Historical Studies
Presented to Erica Reiner. New Haven: American Oriental Society. American Oriental
Series 67. 1 25- 1 38.
Gibson, McGuire and Robert D. Biggs. 1977. Seals and Sealing in the Ancient Near East.
Malibu: Undena. Bibliotheca Mesopotamica 6.
-----. 1987. The Organization of Power: Aspects of Bureacracy in the Ancient Near East.
Chicago: University of Chicago Press. Studies in Ancient Oriental Civilizations 46.
Gordon, Cyrus. 1987. "Eblaitica". in: Cyrus Gordon et al (eds), Eblaitica: Essays on the
Ebla Archives and Eblaite Languag. Winona Lake: Eisenbrauns. Vol. 1 : 19-28.
Gragg, Gene. 1968. "The Syntax of the Copula in Sumerian". in: I.W.M. Verhaar (ed),
The Verb "Be" and its Synonyms. Dordrecht. Vol. 3: 86-109.
----- 1972a. "Sumerian and Selected Afro-Asiatic Languages". in: P. Peranteau (ed), The
Chicago Which Hunt. Chicago: Chicago Linguistic Society. 153- 1 69.
----- 1972b. "Observations on Grammatical Variation in Sumerian Literary Texts".
JAOS 92: 204-21 3.
----- 1973a. Sumerian Dimensional Infixes. Kevelaer: Verlag Butzon & Bercker. Alter
Orient und Altes Testament, Sonderreihe 5.
----- 1973b. "A Class of 'When' Clauses in Sumerian". JNES 32: 124- 134.
----- 1973c. "Linguistics, Method, and Extinct Languages: The Case of Sumerian". Or
42: 78-96.
Grayson, A. Kirk. 1980. "Histories and Historians of the Ancient Near East: Assyria and
Babylonia". Or 49: 140- 194.
Green, M.W. 198 1 . "The Construction and Implementation of the Cuneiform Writing
System". Visible Language 15: 345-372.
Green, M.W. and Hans I. Nissen. 1987. Zeichenliste der archaischen Texte aus Uruk.
Berlin: Gebr. Mann Verlag. Archaische Texte aus Uruk Band 2.
Haayer, G. 1 986. "Languages in Contact: The Case of Akkadian and Sumerian". in:
H.L.I. Vanstiphout et al (eds), Scripta Signa Vocis. Groningen: Egbert Forsten. 77-
84.
Hallo, William W. 1957. Early Mesopotamian Royal Titles: A Philologic and Historical
Analysis. New Haven: American Oriental Society.
1962. "The Royal Inscriptions of Ur: A Typology". HUCA 33: 1-43.
----- 1963. "Lexical Notes on the Neo-Sumerian Metal Industry". BiOr20: 1 36-141.
----- 1966. "The Coronation of Ur-Nammu". JCS 20: 133-141.
Hallo, William W. and 1.1.A. van Dijk. 1968. The Exaltation of Inanna. New Haven:
Yale University Press. Yale Near Eastern Researches 3.
Heimpel, Wolfgang. 1987. "Gudea's Fated Brick". JNES 46: 205-2 1 1 .
Horsnell, Malcolm. 1977. "The Grammar and Syntax of the Year-Names of the First Dy
nasty of Babylon". JNES 36: 277-285.
lacobsen, Thorkild. 1953. "The Reign of Ibbi-Suen". Reprinted in: Moran (1970) 173-
186; 408-41 6.
----- 1956. "Introduction to the Chicago Grammatical Texts". in: MSL IV, 1 *-50*.
Bibliography 295
----- 1957. "Early Political Development in Mesopotamia". Reprinted in: Moran (1970)
1 32- 156; 366-396.
1965. "About the Sumerian Verb". Reprinted in: Moran (1970) 245-270; 430-466.
1968. "Some Sumerian City-Names". JCS 2 1 : 100- 103.
1973. "Notes on the Sumerian Genitive". JNES 32: 161-166.
1974. "Very Ancient Texts: Babylonian Grammatical Texts". in: D. Hymes (ed),
Studies in the History of Linguistics: Traditions and Paradigms. Bloomington:
Indiana University Press. 41-62.
----- 1987. The Harps that Once... Sumerian Poetry in Translation. New Haven: Yale
University Press.
----- 1988. "Sumerian Grammar Today". JADS 108: 123- 1 33.
Kaneva, I.T. 1970. "Participles in Sumerian". MID 16: 541-565.
Kang, S.T. 1968. A Study of the Sumerian Verb in Bilingual Grammatical Texts. Cin-
cinnati: Hebrew Union College [dissertation].
Kienast, Burkhart. 1975. "Zur Wortbildung des Sumerischen". ZA 65: 1-27.
----- 198 1a. "1st das Neusumerische eine lebende Sprache?" CRAI28: 105- 1 1 1 .
----- 198 1 b. "Probleme der sumerischen Grammatik". ZA 70: 1-35.
Kramer, Samuel Noah. 1936. "Studies in Sumerian Phonetics". ArOr 8: 1 8-33.
----- 1940. Lamentation over the Destruction of Ur. Chicago: University of Chicago
Press. AS 1 2.
----- 1967. "The Death of Ur-Nammu and His Descent to the Netherworld". JCS 2 1 :
104-122.
----- 1983. "The Ur-Nammu Law Code: Who Was Its Author?" Or 52: 453-456.
Kraus, ER. 1958. "DLtil.1a. Sumerische Prozessprotokolle und Verwandtes aus der Zeit
der Ill. Dynastie von Ur". BiOI 15: 70-84.
Krecher, Joachim. 1967. "Zum Emesal-Dialekt des Sumerischen". in: Heidelberger Stu
dien zum Alten Orient. Wiesbaden: Otto Harrassowitz. 87- 1 1 O.
----- 1967-68. "Die pluralischen Verba fur 'gehen' und 'stehen' im Sumerischen". WD 4:
1-1 1.
----- 1978. "Die Form und der Gebrauch der nominalen Verbalformen und die Determina
tion im Sumerischen". Dr47: 376-403.
1985. "Die Im/-Prilfixe des sumerischen Verbums". Or 54: 1 33- 1 8 1 .
----- 1987a. "DU KUx (-r) 'eintreten', 'eineinbringen"' . ZA 77: 7-2 1 .
=
Cong@gational Lament. New Haven: Yale University Press. Yale Near Eastern Re
searches 6.
Labat, Rene. 1 976. Manuel d'epigraphie akkadienne5 . Paris: Paul Geuthner.
Lambert, Maurice. 1 957. "Les noms du pere en Sumerien". Proceedings of the Twenty
Second Congress of Orientalists. Leiden: E.J. Brill. Vol. 2: 27-29.
----- 1 972-7 3 . "Probleme essentiel dans l'etude du Sumerien: Les prefixes verbaux".
GLECS 1 7 : 1 3-2 1 ; 97- 1 04.
Lambert, W.G. 1 975. (review of Henri Limet, Les legendes des sceaux cassites [ 1 97 1 ] ) .
BiOr 32: 2 1 9-223.
Larsen, Mogens Trolle. 1987. "The Mesopotamian Lukewarm Mind: Reflections on
Science, Divination, and Literacy". in: Francesca Rochberg-Halton (ed), Lang.!!M
Literature, and History: Philological and Historical S tudies Presented to Erica Reiner.
New Haven: American Oriental Society. American Oriental Series 67. 203-205.
Lieberman, Stephen J. 1 977. The Sumerian Loanwords in Old-Babylonian Akkadian.
Missoula: Scholars Press. Harvard Semitic S tudies 22.
----- 1 978. (letter to editor). Scientific American 239/ 1 0: 1 5 .
----- 1 979. "The phoneme /0/ in Sumerian". in: Studies in Honor of Tom B. Jones.
Kevelaer: Verlag Butzon & Bercker. Alter Orient und Altes Testament 203 . 2 1 -28.
----- 1 980. "Of Clay Pebbles, Hollow Clay Balls, and Writing: A Sumerian View".
AlA 84: 339-358.
Limet, Henri. 1 975a. "Le morpheme suffixe /-a/ en Sumerien". RA 69: 5- 1 8 .
----- 1 975b. "Permanence et changement dans l a toponymie de l a Mesopotamie antique".
in: La Toponymie antique. Actes du Colloque de Strasbourg 1 2- 1 4 juin 1 975. Leiden:
EJ. Brill. 83- 1 1 5 .
----- 1 978. "Etude semantique d e ma.da, kur, kalam". RA 7 2 : 1 - 1 2.
Martiny, Giinther. 1 940. "The Orientation of the Gimilsin Temple and the Palace Chapel".
in: Frankfort, Henri, Seton Lloyd, and Thorkild J acobsen, The Gimilsin Temple and
the Palace of the Rulers at Tell Asmar. Chicago: University of Chicago Press. OIP
43. 92-96.
Michalowski, Piotr. 1 980a. "Sumerian as an Ergative Language". lCS 32: 86- 1 03.
----- 1 980b. "Konigsbriefe". RLA 6: 5 1 -59.
----- 1 987. "Charisma and Control: On Continuity and Change in Early Mesopotamian
Bureacratic S ystems". in: Gibson and Biggs ( 1 987) 50-68.
Moorey, P.R.S . 1984. "Where did they bury the kings of the IIIrd dynasty of Ur?". Iraq
46: 1 - 1 8.
Moran, William (ed). 1 970. Toward the Image of Tammuz. Cambridge: Harvard Univer
sity Press.
Nissen, Hans J. 1 985. "The Emergence of Writing in the Ancient Near East". Inter
disciplinary Science Revie ws 1 0/4: 349-36 1 .
----- 1 986. "The archaic texts from Uruk". World Archaeology 1 7/ 3 : 3 1 7-334.
----- 1988. The Early History of the Ancient Near East: 9000-2000 B .C. Translated by
Elizabeth Lutzeier, with Kenneth J. Northcott. Chicago: University of Chicago Press.
Oberhuber, Karl. 1 982. "Zum 'Passivum' im Sumerischen". in: W. Meid (ed), Sprach-
Bibliography 297
----- 1972. "Ur III Society: Some Unanswered Questions". CRAI 18: 185-189.
----- 1973. "Problems of Translation". Or 42: 158-161.
Steiner, Gerd. 1978. "Der Gegensatz 'eigenes Land' : 'Ausland, Fremdland, Feindland' in
den Vorstellungen des alten Orients". CRAI25: 633-664.
----- 1980. "The Vocalization of the Sumerian Verbal Morpheme /=ed/ and its Signi
ficance". JNES 40: 21-41.
----- 198 1 . "ljamty. und maru als verbale Kategorien im Sumerischen und im Akka
disc hen". RA 75: 1-14.
Steinkeller, Piotr. 1977. "Seal Practice in the Ur III Period". in: Gibson and Biggs
(1977) 41-53.
1979. "Notes on Sumerian Plural Verbs". Or48: 54-67.
1984. "The Sumerian Verb lugx (LUL)". SELl : 5-17.
1987a. (review of Pennsylvania Sumerian DictionID). lNES 46: 55-59.
1987b. "The Administrative and Economic Organization of the Ur III State: The
Core and the Periphery". in: Gibson and Biggs (1987) 19-41.
----- 1988. "On the Identity of the Toponym LV.SU(.A)". lAOS 108: 197-202.
Stola, R. 198 1-82. "Zu den sumerischen Entsprechungen des akkadischen Imperativs in
spaten zweisprachigen Texten". Af0 28: 79-91 .
Strommenger, Eva. 1980. "The Chronological Division of the Archaic Levels of Uruk
Eanna VI to IIIIIl: Past and Present". AlA 84: 479-487.
Sullivan, Brian. 1979. Sumerian and Akkadian Sentence Structure in Old Babylonian
LiterID Bilingual Texts. Cincinnati: Hebrew Union College [dissertation].
Thomsen, Marie-Louise. 1984. The Sumerian Language: An Introduction to its History
and Grammatical Structure. Copenhagen: Akademisk Forlag. Mesopotamia: Copen
hagen Studies in Assyriology 10.
Tigay, Jeffrey H. 1982. The Evolution of the Gilgamesh Epic. Philadelphia: The Univer
sity of Pennsylvania Press.
Vaiman, A.A. 1974. "Uber die Protosumerische Schrift". Acta Antiqua 22: 15-27.
van Driehl, G. 1982. "Tablets from Jebel Aruda". in: G. van Driehl (ed), Zikir umim:
Assyriological Studies Presented to F.R. Kraus. 12-25.
Vanstiphout, Herman L.J. 1985. "On the Verbal Prefix lif in Standard Sumerian". RA
79: 1-15.
Veenhof, K.R. 1986. "Cuneiform Archives. An Introduction". in: K.R. Veenhof (ed),
Cuneiform Archives and Libraries. Leiden: Nederlands Instituut voor het Nabije Oos
ten.
Walker, c.B.F. 1987. Cuneiform. Berkeley/London: University of California Press and
the British Museum.
Weadock, Penelope. 1975. "The giparu at Ur". Iraq 37: 101-128.
Westenholz, Aage. 1985. "An essay on the Sumerian 'Lexical' Texts of the Third Millen
nium". Or 54: 294-298.
Wiseman, D.J. 1960. "The Goddess Lama at Ur". Iraq 22: 166-171.
Woolley, Sir Leonard. 1982. Ur 'of the Chaldees'. Revised and updated by P.R.S.
Moorey. Ithaca: Cornell University Press.
Bibliography 299
Yoshikawa, Mamoru. 1968a. "The Manl and IJamu Aspects in the Sumerian Verbal
System". Or 37: 401 -416.
----- 1968b. "On the Grammatical Function of -E- of the Sumerian Verbal Suffix -E-DEI-
E-DA(M)". JNES 27: 25 1-261.
1974. "The Maru-Conjugation in the Sumerian Verbal System". Or 43: 17-39.
1979. "Verbal Reduplication in Sumerian". ASJ 1 : 99- 1 19.
1980. "On the Alleged 'Ueberhlingende Vokale''' . ASJ2: 1 88- 195.
198 1a. "The Sumerian Verbal Prefix al-". CRAI28: 66-7 1 .
198 1b. "Plural Expressions in Sumerian Verbs". ASJ3: 1 1 1- 124.
Yushu, Gong. 1987. "ErgativiUit und das Sumerisch". Journal of Ancient Civilizations
2: 85- 120.
Zakar, Andras. 197 1 . "Sumerian-Ural-Altaic Affinities". Current Anthropology 12:
2 15-225.
Zettler, Richard. 1977. "The Sargonic Royal Seal: a Consideration of Sealing in Meso
potamia". in: Gibson and Biggs (1977) 33-39.
Zhi, Yang. 1987. "The Name of the City Adab". Journal of Ancient Civilizations 2:
121- 125.
Concordance of texts used
This concordance lists all the texts used in the Lessons. For each text, the following
information is given: number according to Hallo's catalogue (if available); publication of
cuneiform; publication of transliteration and translation in SAKI; publication of translation
in IRSA; location of photograph. The following abbreviations are used:
Cuneiform
BE = Babylonian Expedition of the University of Pennsylvania, Series A, Cuneiform
Texts. Philadelphia.
BE 1 1 896 (Hilprecht)
=
PBS 1 5 1926
=
I R = 1 86 1
UET = Ur Excavations. Texts. London.
UET 1 = 1 928
UET 3 = 1 937
UET 8 = 1 965
UVB = Vorlaufiger Bericht iiber die von der Deutschen Forschungsgemeinschaft in Uruk
Warka unternommenen Ausgrabungen. Berlin.
UVB 10 = 1 939
UVB 1 2/ 1 3 = 1 956
VAS = Vorderasiatische Denkmaler. Berlin.
VAS 1 = 1 907
YOS = Yale Oriental Series. New Haven.
YOS 1 = 1 9 1 5
Secondary Literature
Textual Concordance
Text 1 Ur-Nammu 8
UET 1 , 40
Drawing of ziggurat: Courtesy of British Museum. Reproduced in Woolley
Bibliography 301
( 1 982) 148
Photograph of ziggurat: Courtesy of Jack Finegan. Reproduced in Finegan
( 1 979) 5 1 . Reproduced courtesy of Westview Press
Photograph of stela: Courtesy of University Museum, University of
Pennsylvania
Text 2 Ur-Nammu 7i
CT 2 1 , 3 : 900 1 5
SAKI 1 86d
Photograph: Courtesy of British Museum. Reproduced in RSA, facing page
280
Photographs of stamp seals: Courtesy of University Museum, University of
Pennsylvania
Text 3 Ur-Nammu 9i
3a CT 2 1 , 2: 90004
Photograph: Courtesy of British Museum. Reproduced in Rall 1 06
3b CT 2 1 , 2: 90009
SAKI 1 86b
IRSA IIIA l c
Text 3c Ur-Nammu 7i
CT 2 1 , 3: 90006
SAKI 1 86d
Text 4 Ur-Nammu 3 1
UET 1 , 34
Text 4a Ur-Nammu 1 1
CT 2 1 , 5 : 9000 1
SAKI 1 96e
Text S Ur-Nammu 23i
UET 1 , 46
Text 6 Ur-Nammu 5ii
UET 1 , 4 1 a
IRSA IIIA lj
Text 7a Ur-Nammu 3iii
Hallo, in: Ancient Mesopotamian Art and Selected Texts. New York: The
Pierpont Morgan Library ( 1 976), p. 22
Photograph : Courtesy of The Pierpont Morgan Library. Reproduced in ibid, p.
17
Photograph of canephore figurines: Courtesy of The Pierpont Morgan Library.
Reproduced in ibid, p. 22
Text 7b Ur-Nammu 3ii
CT 2 1 , 4: 90802
SAKI 1 86g
Text 7c Ur-Nammu Si
CT 2 1 , 5 : 90296
302 Appendix 4
SAKI 1 86f
IRSA IIIAlj
Text 8 Ur-Nammu 1 6
BE 1 , 1 2 1
SAKI 1 88k
Text 8a Ur-Nammu 3i
BE 1, 1 22
SAKI 1 87, note h
Text 9 Ur-Nammu l Oi
CT 2 1 , 7: 90000
SAKI 1 86c
IRSA IIIAld
Drawing of ziggurat: Courtesy of British Museum; reproduced in Woolley
(1982) 234
Text 10 Shulgi 52
VET 1 , 55
Photograph of 30-mina weight: Courtesy of British Museum
Text lOa Ur-Nammu lOi
IR I i5
SAKI 1 86c
IRSA IIIAld
Text 1 1 Shulgi 29
CT 5, 2: 122 1 8
SAKI 194x
IRSA IIIA2u
Photograph: Hallo, in Denise Schmandt-Besserat (ed), The Legacy of Sumer.
Malibu: Undena Publications (1976), p. 133
Text l la Ur-Nammu Iv
Boson, Aegyptus 1 5 (1935) 420
IRSA IIIAl b
Text 12 (Ur-Nammu)
UET 8, 4:21
IRSA IIIAlk
Text 12a Shulgi 20i
PBS 15, 42
IRSA IIIA2g
Photograph is of Shulgi 22ii: McCown, Archaeology 5 (1952) 74
Text 1 3a Amar-Sin 2ii
CT 2 1 ,'24: 90034
SAKI 196b
IRSA IIIA3b
Text 1 3b Amar-Sin 2iii
VAS 1 , 26
Bibliography 303
IRSA llIA3b
Text l 3e Amar-Sin 1
BE 1 , 22
SAKI 196a
IRSA IIIA3a
Text 14 Amar-Sin 5iv
Sumer 3 (1947) le, facing p. 236
IRSA llIA3h
Text 14a Vr-Nammu 22i
VET 1 , 45
Text 1 5 Amar-Sin 3i
er 2 1 , 25f: 908 1 1
SAKI 198d
IRSA IIIA3e
Photograph of pedestal: Courtesy of British Museum. VET 1 , 172, pI. V
Text 1 6 Amar-Sin 1 0
BE 1, 21
SAKI 1989
IRSA IIIA3f
Text 16a Shulgi 46
RO 2 (1925) 1 89
IRSA IIIA2f
Text 17 Amar-Sin 1 1
VVB 1 0, pI. 28
IRSA IIIA3d
Photograph: VVB 10, pI. 23a
Text 18: Shu-Sin 6i
CT 2 1 , 28: 90844
SAKI 200b
IRSA IIIA4e
Text 1 8a Shulgi 43
MDP 6, 22
SAKI 194y
Text 19 Shu-Sin 9iii
YOS 1 , 20
IRSA IIIA4d
Text 19a Amar-Sin 17
BIN 2, 17
Text 20 Shu-Sin 3
OIP 14, 43
IRSA IIIC1 b
Text 21a Shu-Sin 17
de Sarzee, Deeouvertes en Chaldee (1912) voI. 2 pI. 26-bis: 5
304 Appendix 4
SAKI 202e
IRSA IIIA4i
Text 21b Ibbi-Sin 7i
UET 3, 52
Photograph of Old Akkadian seal: Courtesy of British Museum.
Text 21c Ibbi-Sin 8iii
PBS 1 3, 5: CBS 12570
Text 21d: Shulgi 5 l iii
UET 1 , 287
Photograph: The Illustrated Bible DictionID.)' (Tyndale, 1980), vo!. 3, 1 635
Text 22 Shulgi 47
CT 2 1 , 9: 89 1 3 1
SAKI 194z
IRSA IIIA2w
Collation: Gadd, Iraq 1 0 (1948) 98 n. 1
Photograph: Courtesy of British Museum. Reproduced in RSA, facing p. 246
Text 22a Ur-Nammu 36
CT 2 1 , 6: 891 26
SAKI 1 88n
IRSA IIIA1m
Photograph: Courtesy of British Museum. Reproduced in RSA, opposite p.
246
Appendix 5
Further Work
The Ur III royal inscriptions studied in this book present only a limited picture of
S umerian. For example, not all the modal-prefixes occur, and there is only limited use of
the first and second persons. In order to deepen one's knowledge of S umerian, it is neces
sary to do two things: read well edited Sumerian texts, and read the most important and/or
recent secondary literature about Sumerian.
More "variety" in grammar occurs in Sumerian literary texts, and for that reason they
are perhaps the most useful texts to read at this stage. However, this is not as easy as it
sounds. Because most literary texts were copied down in the Old Babylonian period or
later, they are often influenced by Akkadian grammar, and may contain forms which are
simply "wrong" by the normative rules of Classical Sumerian grammar; such wrong forms
can be disconcerting to a relati ve beginner.
To obviate this problem, it is necessary to work through literary texts which are well
edited. However, some Sumerologists are less interested in grammatical matters than other
S umerologists, and so may not discuss such matters as, for example, a seemingly incorrect
use of a personal-affix. Some scholars are more interested in lexicographical matters, and
may devote seemingly inordinate amount of space for citing all references for particular
words, instead of focusing on the grammar (given the lack of a complete up-to-date
dictionary of S umerian, this is often necessary).
In general, it is always valuable to first skim through an edition of a literary text, in
order to determine what kind of emphasis the modern-day editor is placing in his
commentary - is it primarily a grammatical commentary, or lexicographic, or stylistic, etc.
When reading through a new text, one will encounter variations of constructions seen
previously, or completely new constructions. These may or may not be discussed in a
commentary. Upon encountering a new form or construction, the first step is to isolate the
problem, that is, determine where in the grammar the problem lies: is it a previously unseen
modal-prefix, or a strange use of a dimensional-prefix, or a problem in a temporal clause,
etc. At this stage, one must turn to the standard grammars, and to other secondary
li terature.
The most general grammars which can be examined are, in chronological order (these
are further discussed below): Poebel ( 1 923); Falkenstein ( 1 959); Falkenstein ( 1 978 2 );
Romer ( 1983 4 ); Thomsen ( 1 984). Thomsen will be of most value to the student; she also
includes references to other secondary literature.
In addition to these more general grammars, there are certain articles, on specific
aspects of the grammar, which are always worth consulting. These are referred to in
Tomsen, but it is useful to keep photocopies of them at hand. Full references are given in
theBibliography.
305
306 Appendix 5
Romer ( 19834 ) is essentially an updated Falkenstein; the section entitled "Einiges zur
Sprache" is less than fifty pages long. However, it has an enormous amount of
bibliographic references - there are 641 footnotes; these references are what make the work
valuable. A new edition is being prepared.
Finally, the most recent grammar is that of Thomsen ( 1984); there is an important
review by Jacobsen (1988). Thomsen bases her work on the Old Babylonian literary texts,
because these exhibit the greatest variety of grammatical variation. The danger with this
approach is the fact that in these texts one must always be on guard against Akkadian
influence. In spite of this methodological criticism, Thomsen contains a balanced,
thorough, and relatively uncontroversial presentation of the grammar, more-or-Iess in the
Falkensteinian tradition. Her book would be difficult for anyone who does not know the
principles of cuneiform script, but for those who already know some Sumerian, it will
serve as a standard reference for some time.
This Manual will be followed by a second volume consisting of extracts from Inanna's
Descent, with a commentary focussing on the grammar of the text.
Appendix 6
Topical Index
The index is designed to tie together information scattered throughout this book, in
order to permit quick reference and review. Only significant discussions are listed. Re
ferences are to pages.
309
3 10 Appendix 6
Modular Preli
minar)' Reports
TPR 1 G. BUCCELLATI
A Cuneiform Tablet from !he Second Season. Syro.lft'.wpowmwn Swdies 1/4 (1977).
TPR J M. KELLY-BUCCELLATI and L. MOUNTWILUAMS
Object Typology of Ihe Second Season (Excluding Vessels ,lnd Lithics)
SyrrrMewpatomiall Sflldif'. 1/5 (1917).
Tl'R s: A. MAHMOUD
Die Induslrie der islamischen KeraOlik ous der lweilen Season
SyrfJ-MeWl/lofOmioll Sfudi,',\ 2/5 (1 978).
TPR 7 O. ROUAULT
Les Documents epigraphiques de la Iroisieme SIIison, Syro-Mf''opotollliall SmdieI 2/7 ( 1979),
Tl'R 8 L. MOUNTWILLiAMS
Objecl Typology of the Third Season: The Third and Second Millennia
S.l'rl).M"SIIfJf!lomiall Sf/uJies 3/2 (1 980).
FinflJ Reports
R I \ 'II I"S \IJ)S ,\NB RISb\RCH TOOLS IN ANClf:NT NEAR I\SUR;\; STlIOll:S